From a362960b29749750168601f3969ec9364082ee59 Mon Sep 17 00:00:00 2001 From: Gustavo Romero Date: Tue, 26 Aug 2025 11:21:27 +0100 Subject: [PATCH 01/32] target/arm: Clean up of register field definitions Clean up the definitions of NSW and NSA fields in the VTCR register. These two fields are already defined properly using FIELD() so they are actually duplications. Also, define the NSW and NSA fields in the VSTCR register using FIELD() and remove their definitions based on VTCR fields. Signed-off-by: Gustavo Romero Message-id: 20250725014755.2122579-1-gustavo.romero@linaro.org Reviewed-by: Peter Maydell Signed-off-by: Peter Maydell --- target/arm/internals.h | 8 +++----- target/arm/ptw.c | 8 ++++---- 2 files changed, 7 insertions(+), 9 deletions(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index 1b3d0244fd..3f86b07044 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -113,11 +113,6 @@ FIELD(DBGWCR, WT, 20, 1) FIELD(DBGWCR, MASK, 24, 5) FIELD(DBGWCR, SSCE, 29, 1) -#define VTCR_NSW (1u << 29) -#define VTCR_NSA (1u << 30) -#define VSTCR_SW VTCR_NSW -#define VSTCR_SA VTCR_NSA - /* Bit definitions for CPACR (AArch32 only) */ FIELD(CPACR, CP10, 20, 2) FIELD(CPACR, CP11, 22, 2) @@ -220,6 +215,9 @@ FIELD(VTCR, NSA, 30, 1) FIELD(VTCR, DS, 32, 1) FIELD(VTCR, SL2, 33, 1) +FIELD(VSTCR, SW, 29, 1) +FIELD(VSTCR, SA, 30, 1) + #define HCRX_ENAS0 (1ULL << 0) #define HCRX_ENALS (1ULL << 1) #define HCRX_ENASR (1ULL << 2) diff --git a/target/arm/ptw.c b/target/arm/ptw.c index 561bf2678e..ed5c728eab 100644 --- a/target/arm/ptw.c +++ b/target/arm/ptw.c @@ -193,9 +193,9 @@ static ARMMMUIdx ptw_idx_for_stage_2(CPUARMState *env, ARMMMUIdx stage2idx) return ARMMMUIdx_Phys_Realm; case ARMSS_Secure: if (stage2idx == ARMMMUIdx_Stage2_S) { - s2walk_secure = !(env->cp15.vstcr_el2 & VSTCR_SW); + s2walk_secure = !(env->cp15.vstcr_el2 & R_VSTCR_SW_MASK); } else { - s2walk_secure = !(env->cp15.vtcr_el2 & VTCR_NSW); + s2walk_secure = !(env->cp15.vtcr_el2 & R_VTCR_NSW_MASK); } return s2walk_secure ? ARMMMUIdx_Phys_S : ARMMMUIdx_Phys_NS; default: @@ -3372,9 +3372,9 @@ static bool get_phys_addr_twostage(CPUARMState *env, S1Translate *ptw, */ if (in_space == ARMSS_Secure) { result->f.attrs.secure = - !(env->cp15.vstcr_el2 & (VSTCR_SA | VSTCR_SW)) + !(env->cp15.vstcr_el2 & (R_VSTCR_SA_MASK | R_VSTCR_SW_MASK)) && (ipa_secure - || !(env->cp15.vtcr_el2 & (VTCR_NSA | VTCR_NSW))); + || !(env->cp15.vtcr_el2 & (R_VTCR_NSA_MASK | R_VTCR_NSW_MASK))); result->f.attrs.space = arm_secure_to_space(result->f.attrs.secure); } From 2561626c25e05bc6f673b29f7dd46a20fa6719c3 Mon Sep 17 00:00:00 2001 From: Pierrick Bouvier Date: Tue, 26 Aug 2025 11:21:27 +0100 Subject: [PATCH 02/32] tests/functional/test_aarch64_device_passthrough: update image TF-A needs to be patched to enable support for FEAT_TCR2 and FEAT_SCTLR2. This new image contains updated firmware. Signed-off-by: Pierrick Bouvier Signed-off-by: Richard Henderson Message-id: 20250727074202.83141-2-richard.henderson@linaro.org Message-ID: <20250719035838.2284029-2-pierrick.bouvier@linaro.org> Signed-off-by: Peter Maydell --- .../aarch64/test_device_passthrough.py | 27 ++++++++++--------- 1 file changed, 14 insertions(+), 13 deletions(-) diff --git a/tests/functional/aarch64/test_device_passthrough.py b/tests/functional/aarch64/test_device_passthrough.py index 1f3f158a9f..17437784bb 100755 --- a/tests/functional/aarch64/test_device_passthrough.py +++ b/tests/functional/aarch64/test_device_passthrough.py @@ -9,7 +9,7 @@ # # SPDX-License-Identifier: GPL-2.0-or-later -import os +from os.path import join from qemu_test import QemuSystemTest, Asset from qemu_test import exec_command, wait_for_console_pattern @@ -77,15 +77,16 @@ echo device_passthrough_test_ok class Aarch64DevicePassthrough(QemuSystemTest): - # https://github.com/pbo-linaro/qemu-linux-stack + # https://github.com/pbo-linaro/qemu-linux-stack/tree/device_passthrough + # $ ./build.sh && ./archive_artifacts.sh out.tar.xz # # Linux kernel is compiled with defconfig + # IOMMUFD + VFIO_DEVICE_CDEV + ARM_SMMU_V3_IOMMUFD # https://docs.kernel.org/driver-api/vfio.html#vfio-device-cde ASSET_DEVICE_PASSTHROUGH_STACK = Asset( - ('https://fileserver.linaro.org/s/fx5DXxBYme8dw2G/' - 'download/device_passthrough.tar.xz'), - '812750b664d61c2986f2b149939ae28cafbd60d53e9c7e4b16e97143845e196d') + ('https://github.com/pbo-linaro/qemu-linux-stack/' + 'releases/download/build/device_passthrough-c3fb84a.tar.xz'), + '15ac2b02bed0c0ea8e3e007de0bcfdaf6fd51c1ba98213f841dc7d01d6f72f04') # This tests the device passthrough implementation, by booting a VM # supporting it with two nvme disks attached, and launching a nested VM @@ -96,16 +97,16 @@ class Aarch64DevicePassthrough(QemuSystemTest): self.vm.set_console() - stack_path_tar_gz = self.ASSET_DEVICE_PASSTHROUGH_STACK.fetch() - self.archive_extract(stack_path_tar_gz, format="tar") + stack_path_tar = self.ASSET_DEVICE_PASSTHROUGH_STACK.fetch() + self.archive_extract(stack_path_tar, format="tar") stack = self.scratch_file('out') - kernel = os.path.join(stack, 'Image.gz') - rootfs_host = os.path.join(stack, 'host.ext4') - disk_vfio = os.path.join(stack, 'disk_vfio') - disk_iommufd = os.path.join(stack, 'disk_iommufd') - guest_cmd = os.path.join(stack, 'guest.sh') - nested_guest_cmd = os.path.join(stack, 'nested_guest.sh') + kernel = join(stack, 'Image.gz') + rootfs_host = join(stack, 'host.ext4') + disk_vfio = join(stack, 'disk_vfio') + disk_iommufd = join(stack, 'disk_iommufd') + guest_cmd = join(stack, 'guest.sh') + nested_guest_cmd = join(stack, 'nested_guest.sh') # we generate two random disks with open(disk_vfio, "wb") as d: d.write(randbytes(512)) with open(disk_iommufd, "wb") as d: d.write(randbytes(1024)) From 84249d026bc3878c196ecd3e8558609ba9260eb6 Mon Sep 17 00:00:00 2001 From: Pierrick Bouvier Date: Tue, 26 Aug 2025 11:21:27 +0100 Subject: [PATCH 03/32] tests/functional/test_aarch64_rme: update image MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit TF-A needs to be patched to enable support for FEAT_TCR2 and FEAT_SCTLR2. This new image contains updated firmware. Signed-off-by: Pierrick Bouvier Signed-off-by: Richard Henderson Reviewed-by: Philippe Mathieu-Daudé Tested-by: Philippe Mathieu-Daudé Message-id: 20250727074202.83141-3-richard.henderson@linaro.org Message-ID: <20250719035838.2284029-3-pierrick.bouvier@linaro.org> [PMM: switch to os.makedirs(..., exist_ok=True) to improve robustness when re-run after test was cancelled midway] Signed-off-by: Peter Maydell --- tests/functional/aarch64/test_rme_sbsaref.py | 64 ++++++++------- tests/functional/aarch64/test_rme_virt.py | 85 +++++++------------- 2 files changed, 66 insertions(+), 83 deletions(-) diff --git a/tests/functional/aarch64/test_rme_sbsaref.py b/tests/functional/aarch64/test_rme_sbsaref.py index 100f1c7738..ca892e0a8c 100755 --- a/tests/functional/aarch64/test_rme_sbsaref.py +++ b/tests/functional/aarch64/test_rme_sbsaref.py @@ -10,21 +10,23 @@ # SPDX-License-Identifier: GPL-2.0-or-later import os +from os.path import join +import shutil from qemu_test import QemuSystemTest, Asset, wait_for_console_pattern from qemu_test import exec_command_and_wait_for_pattern -from test_rme_virt import test_realms_guest class Aarch64RMESbsaRefMachine(QemuSystemTest): - # Stack is built with OP-TEE build environment from those instructions: + # Stack is inspired from: # https://linaro.atlassian.net/wiki/spaces/QEMU/pages/29051027459/ - # https://github.com/pbo-linaro/qemu-rme-stack + # https://github.com/pbo-linaro/qemu-linux-stack/tree/rme_sbsa_release + # ./build.sh && ./archive_artifacts.sh out.tar.xz ASSET_RME_STACK_SBSA = Asset( - ('https://fileserver.linaro.org/s/KJyeBxL82mz2r7F/' - 'download/rme-stack-op-tee-4.2.0-cca-v4-sbsa.tar.gz'), - 'dd9ab28ec869bdf3b5376116cb3689103b43433fd5c4bca0f4a8d8b3c104999e') + ('https://github.com/pbo-linaro/qemu-linux-stack/' + 'releases/download/build/rme_sbsa_release-a7f02cf.tar.xz'), + '27d8400b11befb828d6db0cab97e7ae102d0992c928d3dfbf38b24b6cf6c324c') # This tests the FEAT_RME cpu implementation, by booting a VM supporting it, # and launching a nested VM using it. @@ -35,35 +37,41 @@ class Aarch64RMESbsaRefMachine(QemuSystemTest): self.vm.set_console() - stack_path_tar_gz = self.ASSET_RME_STACK_SBSA.fetch() - self.archive_extract(stack_path_tar_gz, format="tar") + stack_path_tar = self.ASSET_RME_STACK_SBSA.fetch() + self.archive_extract(stack_path_tar, format="tar") - rme_stack = self.scratch_file('rme-stack-op-tee-4.2.0-cca-v4-sbsa') - pflash0 = os.path.join(rme_stack, 'images', 'SBSA_FLASH0.fd') - pflash1 = os.path.join(rme_stack, 'images', 'SBSA_FLASH1.fd') - virtual = os.path.join(rme_stack, 'images', 'disks', 'virtual') - drive = os.path.join(rme_stack, 'out-br', 'images', 'rootfs.ext4') + rme_stack = self.scratch_file('.') + pflash0 = join(rme_stack, 'out', 'SBSA_FLASH0.fd') + pflash1 = join(rme_stack, 'out', 'SBSA_FLASH1.fd') + rootfs = join(rme_stack, 'out', 'host.ext4') - self.vm.add_args('-cpu', 'max,x-rme=on,pauth-impdef=on') + efi = join(rme_stack, 'out', 'EFI') + os.makedirs(efi, exist_ok=True) + shutil.copyfile(join(rme_stack, 'out', 'Image'), join(efi, 'Image')) + with open(join(efi, 'startup.nsh'), 'w') as startup: + startup.write('fs0:Image nokaslr root=/dev/vda rw init=/init --' + ' /host/out/lkvm run --realm' + ' -m 256m' + ' --restricted_mem' + ' --kernel /host/out/Image' + ' --disk /host/out/guest.ext4' + ' --params "root=/dev/vda rw init=/init"') + + self.vm.add_args('-cpu', 'max,x-rme=on') + self.vm.add_args('-smp', '2') self.vm.add_args('-m', '2G') self.vm.add_args('-M', 'sbsa-ref') self.vm.add_args('-drive', f'file={pflash0},format=raw,if=pflash') self.vm.add_args('-drive', f'file={pflash1},format=raw,if=pflash') - self.vm.add_args('-drive', f'file=fat:rw:{virtual},format=raw') - self.vm.add_args('-drive', f'format=raw,if=none,file={drive},id=hd0') - self.vm.add_args('-device', 'virtio-blk-pci,drive=hd0') - self.vm.add_args('-device', 'virtio-9p-pci,fsdev=shr0,mount_tag=shr0') - self.vm.add_args('-fsdev', f'local,security_model=none,path={rme_stack},id=shr0') - self.vm.add_args('-device', 'virtio-net-pci,netdev=net0') - self.vm.add_args('-netdev', 'user,id=net0') - + self.vm.add_args('-drive', f'file=fat:rw:{efi},format=raw') + self.vm.add_args('-drive', f'format=raw,file={rootfs},if=virtio') + self.vm.add_args('-virtfs', + f'local,path={rme_stack}/,mount_tag=host,' + 'security_model=mapped,readonly=off') self.vm.launch() - # Wait for host VM boot to complete. - wait_for_console_pattern(self, 'Welcome to Buildroot', - failure_message='Synchronous Exception at') - exec_command_and_wait_for_pattern(self, 'root', '#') - - test_realms_guest(self) + # Wait for host and guest VM boot to complete. + wait_for_console_pattern(self, 'root@guest', + failure_message='Kernel panic') if __name__ == '__main__': QemuSystemTest.main() diff --git a/tests/functional/aarch64/test_rme_virt.py b/tests/functional/aarch64/test_rme_virt.py index 8452d27928..bb603aaa26 100755 --- a/tests/functional/aarch64/test_rme_virt.py +++ b/tests/functional/aarch64/test_rme_virt.py @@ -9,50 +9,22 @@ # # SPDX-License-Identifier: GPL-2.0-or-later -import os +from os.path import join from qemu_test import QemuSystemTest, Asset from qemu_test import exec_command, wait_for_console_pattern from qemu_test import exec_command_and_wait_for_pattern -def test_realms_guest(test_rme_instance): - - # Boot the (nested) guest VM - exec_command(test_rme_instance, - 'qemu-system-aarch64 -M virt,gic-version=3 ' - '-cpu host -enable-kvm -m 512M ' - '-M confidential-guest-support=rme0 ' - '-object rme-guest,id=rme0 ' - '-device virtio-net-pci,netdev=net0,romfile= ' - '-netdev user,id=net0 ' - '-kernel /mnt/out/bin/Image ' - '-initrd /mnt/out-br/images/rootfs.cpio ' - '-serial stdio') - # Detect Realm activation during (nested) guest boot. - wait_for_console_pattern(test_rme_instance, - 'SMC_RMI_REALM_ACTIVATE') - # Wait for (nested) guest boot to complete. - wait_for_console_pattern(test_rme_instance, - 'Welcome to Buildroot') - exec_command_and_wait_for_pattern(test_rme_instance, 'root', '#') - # query (nested) guest cca report - exec_command(test_rme_instance, 'cca-workload-attestation report') - wait_for_console_pattern(test_rme_instance, - '"cca-platform-hash-algo-id": "sha-256"') - wait_for_console_pattern(test_rme_instance, - '"cca-realm-hash-algo-id": "sha-512"') - wait_for_console_pattern(test_rme_instance, - '"cca-realm-public-key-hash-algo-id": "sha-256"') - class Aarch64RMEVirtMachine(QemuSystemTest): - # Stack is built with OP-TEE build environment from those instructions: + # Stack is inspired from: # https://linaro.atlassian.net/wiki/spaces/QEMU/pages/29051027459/ - # https://github.com/pbo-linaro/qemu-rme-stack + # https://github.com/pbo-linaro/qemu-linux-stack/tree/rme_release + # ./build.sh && ./archive_artifacts.sh out.tar.xz ASSET_RME_STACK_VIRT = Asset( - ('https://fileserver.linaro.org/s/iaRsNDJp2CXHMSJ/' - 'download/rme-stack-op-tee-4.2.0-cca-v4-qemu_v8.tar.gz'), - '1851adc232b094384d8b879b9a2cfff07ef3d6205032b85e9b3a4a9ae6b0b7ad') + ('https://github.com/pbo-linaro/qemu-linux-stack/' + 'releases/download/build/rme_release-86101e5.tar.xz'), + 'e42fef8439badb52a071ac446fc33cff4cb7d61314c7a28fdbe61a11e1faad3a') # This tests the FEAT_RME cpu implementation, by booting a VM supporting it, # and launching a nested VM using it. @@ -63,15 +35,16 @@ class Aarch64RMEVirtMachine(QemuSystemTest): self.vm.set_console() - stack_path_tar_gz = self.ASSET_RME_STACK_VIRT.fetch() - self.archive_extract(stack_path_tar_gz, format="tar") + stack_path_tar = self.ASSET_RME_STACK_VIRT.fetch() + self.archive_extract(stack_path_tar, format="tar") - rme_stack = self.scratch_file('rme-stack-op-tee-4.2.0-cca-v4-qemu_v8') - kernel = os.path.join(rme_stack, 'out', 'bin', 'Image') - bios = os.path.join(rme_stack, 'out', 'bin', 'flash.bin') - drive = os.path.join(rme_stack, 'out-br', 'images', 'rootfs.ext4') + rme_stack = self.scratch_file('.') + kernel = join(rme_stack, 'out', 'Image') + bios = join(rme_stack, 'out', 'flash.bin') + rootfs = join(rme_stack, 'out', 'host.ext4') - self.vm.add_args('-cpu', 'max,x-rme=on,pauth-impdef=on') + self.vm.add_args('-cpu', 'max,x-rme=on') + self.vm.add_args('-smp', '2') self.vm.add_args('-m', '2G') self.vm.add_args('-M', 'virt,acpi=off,' 'virtualization=on,' @@ -79,23 +52,25 @@ class Aarch64RMEVirtMachine(QemuSystemTest): 'gic-version=3') self.vm.add_args('-bios', bios) self.vm.add_args('-kernel', kernel) - self.vm.add_args('-drive', f'format=raw,if=none,file={drive},id=hd0') - self.vm.add_args('-device', 'virtio-blk-pci,drive=hd0') - self.vm.add_args('-device', 'virtio-9p-device,fsdev=shr0,mount_tag=shr0') - self.vm.add_args('-fsdev', f'local,security_model=none,path={rme_stack},id=shr0') - self.vm.add_args('-device', 'virtio-net-pci,netdev=net0') - self.vm.add_args('-netdev', 'user,id=net0') + self.vm.add_args('-drive', f'format=raw,file={rootfs},if=virtio') + self.vm.add_args('-virtfs', + f'local,path={rme_stack}/,mount_tag=host,' + 'security_model=mapped,readonly=off') # We need to add nokaslr to avoid triggering this sporadic bug: # https://gitlab.com/qemu-project/qemu/-/issues/2823 - self.vm.add_args('-append', 'root=/dev/vda nokaslr') + self.vm.add_args('-append', + 'nokaslr root=/dev/vda rw init=/init --' + ' /host/out/lkvm run --realm' + ' -m 256m' + ' --restricted_mem' + ' --kernel /host/out/Image' + ' --disk /host/out/guest.ext4' + ' --params "root=/dev/vda rw init=/init"') self.vm.launch() - # Wait for host VM boot to complete. - wait_for_console_pattern(self, 'Welcome to Buildroot', - failure_message='Synchronous Exception at') - exec_command_and_wait_for_pattern(self, 'root', '#') - - test_realms_guest(self) + # Wait for host and guest VM boot to complete. + wait_for_console_pattern(self, 'root@guest', + failure_message='Kernel panic') if __name__ == '__main__': QemuSystemTest.main() From 8a60ffe9a8f46ed514656eb4a40d1386c439daf8 Mon Sep 17 00:00:00 2001 From: Gustavo Romero Date: Tue, 26 Aug 2025 11:21:28 +0100 Subject: [PATCH 04/32] target/arm: Implement FEAT_SCTLR2 and enable with -cpu max Add FEAT_SCTLR2, which introduces the SCTLR2_EL1, SCTLR2_EL2, and SCTLR2_EL3 registers. These registers are extensions of the SCTLR_ELx ones. Signed-off-by: Gustavo Romero Reviewed-by: Richard Henderson Signed-off-by: Richard Henderson Message-id: 20250727074202.83141-4-richard.henderson@linaro.org Message-ID: <20250711140828.1714666-4-gustavo.romero@linaro.org> [rth: Remove FEAT_MEC code; handle SCR and HCRX enable bits.] Signed-off-by: Richard Henderson Signed-off-by: Peter Maydell --- docs/system/arm/emulation.rst | 1 + target/arm/cpu-features.h | 5 ++ target/arm/cpu.c | 3 ++ target/arm/cpu.h | 15 ++++++ target/arm/helper.c | 97 ++++++++++++++++++++++++++++++++--- target/arm/internals.h | 1 + target/arm/tcg/cpu64.c | 5 +- 7 files changed, 119 insertions(+), 8 deletions(-) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index 890dc6fee2..66043b0747 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -121,6 +121,7 @@ the following architecture extensions: - FEAT_RPRES (Increased precision of FRECPE and FRSQRTE) - FEAT_S2FWB (Stage 2 forced Write-Back) - FEAT_SB (Speculation Barrier) +- FEAT_SCTLR2 (Extension to SCTLR_ELx) - FEAT_SEL2 (Secure EL2) - FEAT_SHA1 (SHA1 instructions) - FEAT_SHA256 (SHA256 instructions) diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index 5876162428..e372543bf3 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -904,6 +904,11 @@ static inline bool isar_feature_aa64_nv2(const ARMISARegisters *id) return FIELD_EX64_IDREG(id, ID_AA64MMFR2, NV) >= 2; } +static inline bool isar_feature_aa64_sctlr2(const ARMISARegisters *id) +{ + return FIELD_EX64_IDREG(id, ID_AA64MMFR3, SCTLRX) != 0; +} + static inline bool isar_feature_aa64_pmuv3p1(const ARMISARegisters *id) { return FIELD_EX64_IDREG(id, ID_AA64DFR0, PMUVER) >= 4 && diff --git a/target/arm/cpu.c b/target/arm/cpu.c index e2b2337399..2ab04cb5f7 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -644,6 +644,9 @@ void arm_emulate_firmware_reset(CPUState *cpustate, int target_el) if (cpu_isar_feature(aa64_fgt, cpu)) { env->cp15.scr_el3 |= SCR_FGTEN; } + if (cpu_isar_feature(aa64_sctlr2, cpu)) { + env->cp15.scr_el3 |= SCR_SCTLR2EN; + } } if (target_el == 2) { diff --git a/target/arm/cpu.h b/target/arm/cpu.h index dc9b6dce4c..08a29802e1 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -337,6 +337,7 @@ typedef struct CPUArchState { }; uint64_t sctlr_el[4]; }; + uint64_t sctlr2_el[4]; /* Extension to System control register. */ uint64_t vsctlr; /* Virtualization System control register. */ uint64_t cpacr_el1; /* Architectural feature access control register */ uint64_t cptr_el[4]; /* ARMv8 feature trap registers */ @@ -1420,6 +1421,19 @@ void pmu_init(ARMCPU *cpu); #define SCTLR_SPINTMASK (1ULL << 62) /* FEAT_NMI */ #define SCTLR_TIDCP (1ULL << 63) /* FEAT_TIDCP1 */ +#define SCTLR2_EMEC (1ULL << 1) /* FEAT_MEC */ +#define SCTLR2_NMEA (1ULL << 2) /* FEAT_DoubleFault2 */ +#define SCTLR2_ENADERR (1ULL << 3) /* FEAT_ADERR */ +#define SCTLR2_ENANERR (1ULL << 4) /* FEAT_ANERR */ +#define SCTLR2_EASE (1ULL << 5) /* FEAT_DoubleFault2 */ +#define SCTLR2_ENIDCP128 (1ULL << 6) /* FEAT_SYSREG128 */ +#define SCTLR2_ENPACM (1ULL << 7) /* FEAT_PAuth_LR */ +#define SCTLR2_ENPACM0 (1ULL << 8) /* FEAT_PAuth_LR */ +#define SCTLR2_CPTA (1ULL << 9) /* FEAT_CPA2 */ +#define SCTLR2_CPTA0 (1ULL << 10) /* FEAT_CPA2 */ +#define SCTLR2_CPTM (1ULL << 11) /* FEAT_CPA2 */ +#define SCTLR2_CPTM0 (1ULL << 12) /* FEAT_CAP2 */ + #define CPSR_M (0x1fU) #define CPSR_T (1U << 5) #define CPSR_F (1U << 6) @@ -1712,6 +1726,7 @@ static inline void xpsr_write(CPUARMState *env, uint32_t val, uint32_t mask) #define SCR_HXEN (1ULL << 38) #define SCR_TRNDR (1ULL << 40) #define SCR_ENTP2 (1ULL << 41) +#define SCR_SCTLR2EN (1ULL << 44) #define SCR_GPF (1ULL << 48) #define SCR_NSE (1ULL << 62) diff --git a/target/arm/helper.c b/target/arm/helper.c index 0c1299ff84..11ddeabb13 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -741,6 +741,9 @@ static void scr_write(CPUARMState *env, const ARMCPRegInfo *ri, uint64_t value) if (cpu_isar_feature(aa64_ecv, cpu)) { valid_mask |= SCR_ECVEN; } + if (cpu_isar_feature(aa64_sctlr2, cpu)) { + valid_mask |= SCR_SCTLR2EN; + } } else { valid_mask &= ~(SCR_RW | SCR_ST); if (cpu_isar_feature(aa32_ras, cpu)) { @@ -3907,23 +3910,21 @@ static void hcrx_write(CPUARMState *env, const ARMCPRegInfo *ri, ARMCPU *cpu = env_archcpu(env); uint64_t valid_mask = 0; - /* FEAT_MOPS adds MSCEn and MCE2 */ if (cpu_isar_feature(aa64_mops, cpu)) { valid_mask |= HCRX_MSCEN | HCRX_MCE2; } - - /* FEAT_NMI adds TALLINT, VINMI and VFNMI */ if (cpu_isar_feature(aa64_nmi, cpu)) { valid_mask |= HCRX_TALLINT | HCRX_VINMI | HCRX_VFNMI; } - /* FEAT_CMOW adds CMOW */ if (cpu_isar_feature(aa64_cmow, cpu)) { valid_mask |= HCRX_CMOW; } - /* FEAT_XS adds FGTnXS, FnXS */ if (cpu_isar_feature(aa64_xs, cpu)) { valid_mask |= HCRX_FGTNXS | HCRX_FNXS; } + if (cpu_isar_feature(aa64_sctlr2, cpu)) { + valid_mask |= HCRX_SCTLR2EN; + } /* Clear RES0 bits. */ env->cp15.hcrx_el2 = value & valid_mask; @@ -3981,11 +3982,16 @@ uint64_t arm_hcrx_el2_eff(CPUARMState *env) * This may need to be revisited for future bits. */ if (!arm_is_el2_enabled(env)) { + ARMCPU *cpu = env_archcpu(env); uint64_t hcrx = 0; - if (cpu_isar_feature(aa64_mops, env_archcpu(env))) { - /* MSCEn behaves as 1 if EL2 is not enabled */ + + /* Bits which whose effective value is 1 if el2 not enabled. */ + if (cpu_isar_feature(aa64_mops, cpu)) { hcrx |= HCRX_MSCEN; } + if (cpu_isar_feature(aa64_sctlr2, cpu)) { + hcrx |= HCRX_SCTLR2EN; + } return hcrx; } if (arm_feature(env, ARM_FEATURE_EL3) && !(env->cp15.scr_el3 & SCR_HXEN)) { @@ -4513,6 +4519,8 @@ static void define_arm_vh_e2h_redirects_aliases(ARMCPU *cpu) static const struct E2HAlias aliases[] = { { K(3, 0, 1, 0, 0), K(3, 4, 1, 0, 0), K(3, 5, 1, 0, 0), "SCTLR", "SCTLR_EL2", "SCTLR_EL12" }, + { K(3, 0, 1, 0, 3), K(3, 4, 1, 0, 3), K(3, 5, 1, 0, 3), + "SCTLR2_EL1", "SCTLR2_EL2", "SCTLR2_EL12", isar_feature_aa64_sctlr2 }, { K(3, 0, 1, 0, 2), K(3, 4, 1, 1, 2), K(3, 5, 1, 0, 2), "CPACR", "CPTR_EL2", "CPACR_EL12" }, { K(3, 0, 2, 0, 0), K(3, 4, 2, 0, 0), K(3, 5, 2, 0, 0), @@ -5994,6 +6002,77 @@ static const ARMCPRegInfo actlr2_hactlr2_reginfo[] = { .resetvalue = 0 }, }; +static CPAccessResult sctlr2_el2_access(CPUARMState *env, + const ARMCPRegInfo *ri, + bool isread) +{ + if (arm_current_el(env) < 3 + && arm_feature(env, ARM_FEATURE_EL3) + && !(env->cp15.scr_el3 & SCR_SCTLR2EN)) { + return CP_ACCESS_TRAP_EL3; + } + return CP_ACCESS_OK; +} + +static CPAccessResult sctlr2_el1_access(CPUARMState *env, + const ARMCPRegInfo *ri, + bool isread) +{ + CPAccessResult ret = access_tvm_trvm(env, ri, isread); + if (ret != CP_ACCESS_OK) { + return ret; + } + if (arm_current_el(env) < 2 && !(arm_hcrx_el2_eff(env) & HCRX_SCTLR2EN)) { + return CP_ACCESS_TRAP_EL2; + } + return sctlr2_el2_access(env, ri, isread); +} + +static void sctlr2_el1_write(CPUARMState *env, const ARMCPRegInfo *ri, + uint64_t value) +{ + uint64_t valid_mask = 0; + + value &= valid_mask; + raw_write(env, ri, value); +} + +static void sctlr2_el2_write(CPUARMState *env, const ARMCPRegInfo *ri, + uint64_t value) +{ + uint64_t valid_mask = 0; + + value &= valid_mask; + raw_write(env, ri, value); +} + +static void sctlr2_el3_write(CPUARMState *env, const ARMCPRegInfo *ri, + uint64_t value) +{ + uint64_t valid_mask = 0; + + value &= valid_mask; + raw_write(env, ri, value); +} + +static const ARMCPRegInfo sctlr2_reginfo[] = { + { .name = "SCTLR2_EL1", .state = ARM_CP_STATE_AA64, + .opc0 = 3, .opc1 = 0, .opc2 = 3, .crn = 1, .crm = 0, + .access = PL1_RW, .accessfn = sctlr2_el1_access, + .writefn = sctlr2_el1_write, .fgt = FGT_SCTLR_EL1, + .nv2_redirect_offset = 0x278 | NV2_REDIR_NV1, + .fieldoffset = offsetof(CPUARMState, cp15.sctlr2_el[1]) }, + { .name = "SCTLR2_EL2", .state = ARM_CP_STATE_AA64, + .opc0 = 3, .opc1 = 4, .opc2 = 3, .crn = 1, .crm = 0, + .access = PL2_RW, .accessfn = sctlr2_el2_access, + .writefn = sctlr2_el2_write, + .fieldoffset = offsetof(CPUARMState, cp15.sctlr2_el[2]) }, + { .name = "SCTLR2_EL3", .state = ARM_CP_STATE_AA64, + .opc0 = 3, .opc1 = 6, .opc2 = 3, .crn = 1, .crm = 0, + .access = PL3_RW, .writefn = sctlr2_el3_write, + .fieldoffset = offsetof(CPUARMState, cp15.sctlr2_el[3]) }, +}; + void register_cp_regs_for_features(ARMCPU *cpu) { /* Register all the coprocessor registers based on feature bits */ @@ -7223,6 +7302,10 @@ void register_cp_regs_for_features(ARMCPU *cpu) define_arm_cp_regs(cpu, nmi_reginfo); } + if (cpu_isar_feature(aa64_sctlr2, cpu)) { + define_arm_cp_regs(cpu, sctlr2_reginfo); + } + if (cpu_isar_feature(any_predinv, cpu)) { define_arm_cp_regs(cpu, predinv_reginfo); } diff --git a/target/arm/internals.h b/target/arm/internals.h index 3f86b07044..fb72236255 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -230,6 +230,7 @@ FIELD(VSTCR, SA, 30, 1) #define HCRX_CMOW (1ULL << 9) #define HCRX_MCE2 (1ULL << 10) #define HCRX_MSCEN (1ULL << 11) +#define HCRX_SCTLR2EN (1ULL << 15) #define HPFAR_NS (1ULL << 63) diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c index 35cddbafa4..f4efff03a5 100644 --- a/target/arm/tcg/cpu64.c +++ b/target/arm/tcg/cpu64.c @@ -1247,7 +1247,10 @@ void aarch64_max_tcg_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64MMFR2, E0PD, 1); /* FEAT_E0PD */ SET_IDREG(isar, ID_AA64MMFR2, t); - FIELD_DP64_IDREG(isar, ID_AA64MMFR3, SPEC_FPACC, 1); /* FEAT_FPACC_SPEC */ + t = GET_IDREG(isar, ID_AA64MMFR3); + t = FIELD_DP64(t, ID_AA64MMFR3, SCTLRX, 1); /* FEAT_SCTLR2 */ + t = FIELD_DP64(t, ID_AA64MMFR3, SPEC_FPACC, 1); /* FEAT_FPACC_SPEC */ + SET_IDREG(isar, ID_AA64MMFR3, t); t = GET_IDREG(isar, ID_AA64ZFR0); t = FIELD_DP64(t, ID_AA64ZFR0, SVEVER, 2); /* FEAT_SVE2p1 */ From 6e6d5fb4b928f09b46d0fa99830d75cddb22f73b Mon Sep 17 00:00:00 2001 From: Gustavo Romero Date: Tue, 26 Aug 2025 11:21:28 +0100 Subject: [PATCH 05/32] target/arm: Implement FEAT_TCR2 and enable with -cpu max Add FEAT_TCR2, which introduces the TCR2_EL1 and TCR2_EL2 registers. These registers are extensions of the TCR_ELx registers and provide top-level control of the EL10 and EL20 translation regimes. Signed-off-by: Gustavo Romero Signed-off-by: Richard Henderson Reviewed-by: Richard Henderson Message-id: 20250727074202.83141-5-richard.henderson@linaro.org Message-ID: <20250711140828.1714666-5-gustavo.romero@linaro.org> Reviewed-by: Richard Henderson [rth: Remove FEAT_MEC code; handle SCR and HCRX enable bits.] Signed-off-by: Richard Henderson Signed-off-by: Peter Maydell --- docs/system/arm/emulation.rst | 1 + target/arm/cpu-features.h | 5 +++ target/arm/cpu.c | 3 ++ target/arm/cpu.h | 2 + target/arm/helper.c | 71 +++++++++++++++++++++++++++++++++++ target/arm/internals.h | 19 ++++++++++ target/arm/tcg/cpu64.c | 1 + 7 files changed, 102 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index 66043b0747..1c597d8673 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -149,6 +149,7 @@ the following architecture extensions: - FEAT_SPECRES (Speculation restriction instructions) - FEAT_SSBS (Speculative Store Bypass Safe) - FEAT_SSBS2 (MRS and MSR instructions for SSBS version 2) +- FEAT_TCR2 (Support for TCR2_ELx) - FEAT_TGran16K (Support for 16KB memory translation granule size at stage 1) - FEAT_TGran4K (Support for 4KB memory translation granule size at stage 1) - FEAT_TGran64K (Support for 64KB memory translation granule size at stage 1) diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index e372543bf3..8ec8c3feb3 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -904,6 +904,11 @@ static inline bool isar_feature_aa64_nv2(const ARMISARegisters *id) return FIELD_EX64_IDREG(id, ID_AA64MMFR2, NV) >= 2; } +static inline bool isar_feature_aa64_tcr2(const ARMISARegisters *id) +{ + return FIELD_EX64_IDREG(id, ID_AA64MMFR3, TCRX) != 0; +} + static inline bool isar_feature_aa64_sctlr2(const ARMISARegisters *id) { return FIELD_EX64_IDREG(id, ID_AA64MMFR3, SCTLRX) != 0; diff --git a/target/arm/cpu.c b/target/arm/cpu.c index 2ab04cb5f7..27a4610da5 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -644,6 +644,9 @@ void arm_emulate_firmware_reset(CPUState *cpustate, int target_el) if (cpu_isar_feature(aa64_fgt, cpu)) { env->cp15.scr_el3 |= SCR_FGTEN; } + if (cpu_isar_feature(aa64_tcr2, cpu)) { + env->cp15.scr_el3 |= SCR_TCR2EN; + } if (cpu_isar_feature(aa64_sctlr2, cpu)) { env->cp15.scr_el3 |= SCR_SCTLR2EN; } diff --git a/target/arm/cpu.h b/target/arm/cpu.h index 08a29802e1..c15d79a106 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -366,6 +366,7 @@ typedef struct CPUArchState { uint64_t vsttbr_el2; /* Secure Virtualization Translation Table. */ /* MMU translation table base control. */ uint64_t tcr_el[4]; + uint64_t tcr2_el[3]; uint64_t vtcr_el2; /* Virtualization Translation Control. */ uint64_t vstcr_el2; /* Secure Virtualization Translation Control. */ uint32_t c2_data; /* MPU data cacheable bits. */ @@ -1726,6 +1727,7 @@ static inline void xpsr_write(CPUARMState *env, uint32_t val, uint32_t mask) #define SCR_HXEN (1ULL << 38) #define SCR_TRNDR (1ULL << 40) #define SCR_ENTP2 (1ULL << 41) +#define SCR_TCR2EN (1ULL << 43) #define SCR_SCTLR2EN (1ULL << 44) #define SCR_GPF (1ULL << 48) #define SCR_NSE (1ULL << 62) diff --git a/target/arm/helper.c b/target/arm/helper.c index 11ddeabb13..5a219703ae 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -741,6 +741,9 @@ static void scr_write(CPUARMState *env, const ARMCPRegInfo *ri, uint64_t value) if (cpu_isar_feature(aa64_ecv, cpu)) { valid_mask |= SCR_ECVEN; } + if (cpu_isar_feature(aa64_tcr2, cpu)) { + valid_mask |= SCR_TCR2EN; + } if (cpu_isar_feature(aa64_sctlr2, cpu)) { valid_mask |= SCR_SCTLR2EN; } @@ -3922,6 +3925,9 @@ static void hcrx_write(CPUARMState *env, const ARMCPRegInfo *ri, if (cpu_isar_feature(aa64_xs, cpu)) { valid_mask |= HCRX_FGTNXS | HCRX_FNXS; } + if (cpu_isar_feature(aa64_tcr2, cpu)) { + valid_mask |= HCRX_TCR2EN; + } if (cpu_isar_feature(aa64_sctlr2, cpu)) { valid_mask |= HCRX_SCTLR2EN; } @@ -3989,6 +3995,9 @@ uint64_t arm_hcrx_el2_eff(CPUARMState *env) if (cpu_isar_feature(aa64_mops, cpu)) { hcrx |= HCRX_MSCEN; } + if (cpu_isar_feature(aa64_tcr2, cpu)) { + hcrx |= HCRX_TCR2EN; + } if (cpu_isar_feature(aa64_sctlr2, cpu)) { hcrx |= HCRX_SCTLR2EN; } @@ -4529,6 +4538,8 @@ static void define_arm_vh_e2h_redirects_aliases(ARMCPU *cpu) "TTBR1_EL1", "TTBR1_EL2", "TTBR1_EL12" }, { K(3, 0, 2, 0, 2), K(3, 4, 2, 0, 2), K(3, 5, 2, 0, 2), "TCR_EL1", "TCR_EL2", "TCR_EL12" }, + { K(3, 0, 2, 0, 3), K(3, 4, 2, 0, 3), K(3, 5, 2, 0, 3), + "TCR2_EL1", "TCR2_EL2", "TCR2_EL12", isar_feature_aa64_tcr2 }, { K(3, 0, 4, 0, 0), K(3, 4, 4, 0, 0), K(3, 5, 4, 0, 0), "SPSR_EL1", "SPSR_EL2", "SPSR_EL12" }, { K(3, 0, 4, 0, 1), K(3, 4, 4, 0, 1), K(3, 5, 4, 0, 1), @@ -6073,6 +6084,62 @@ static const ARMCPRegInfo sctlr2_reginfo[] = { .fieldoffset = offsetof(CPUARMState, cp15.sctlr2_el[3]) }, }; +static CPAccessResult tcr2_el2_access(CPUARMState *env, const ARMCPRegInfo *ri, + bool isread) +{ + if (arm_current_el(env) < 3 + && arm_feature(env, ARM_FEATURE_EL3) + && !(env->cp15.scr_el3 & SCR_TCR2EN)) { + return CP_ACCESS_TRAP_EL3; + } + return CP_ACCESS_OK; +} + +static CPAccessResult tcr2_el1_access(CPUARMState *env, const ARMCPRegInfo *ri, + bool isread) +{ + CPAccessResult ret = access_tvm_trvm(env, ri, isread); + if (ret != CP_ACCESS_OK) { + return ret; + } + if (arm_current_el(env) < 2 && !(arm_hcrx_el2_eff(env) & HCRX_TCR2EN)) { + return CP_ACCESS_TRAP_EL2; + } + return tcr2_el2_access(env, ri, isread); +} + +static void tcr2_el1_write(CPUARMState *env, const ARMCPRegInfo *ri, + uint64_t value) +{ + uint64_t valid_mask = 0; + + value &= valid_mask; + raw_write(env, ri, value); +} + +static void tcr2_el2_write(CPUARMState *env, const ARMCPRegInfo *ri, + uint64_t value) +{ + uint64_t valid_mask = 0; + + value &= valid_mask; + raw_write(env, ri, value); +} + +static const ARMCPRegInfo tcr2_reginfo[] = { + { .name = "TCR2_EL1", .state = ARM_CP_STATE_AA64, + .opc0 = 3, .opc1 = 0, .opc2 = 3, .crn = 2, .crm = 0, + .access = PL1_RW, .accessfn = tcr2_el1_access, + .writefn = tcr2_el1_write, .fgt = FGT_TCR_EL1, + .nv2_redirect_offset = 0x270 | NV2_REDIR_NV1, + .fieldoffset = offsetof(CPUARMState, cp15.tcr2_el[1]) }, + { .name = "TCR2_EL2", .state = ARM_CP_STATE_AA64, + .opc0 = 3, .opc1 = 4, .opc2 = 3, .crn = 2, .crm = 0, + .access = PL2_RW, .accessfn = tcr2_el2_access, + .writefn = tcr2_el2_write, + .fieldoffset = offsetof(CPUARMState, cp15.tcr2_el[2]) }, +}; + void register_cp_regs_for_features(ARMCPU *cpu) { /* Register all the coprocessor registers based on feature bits */ @@ -7306,6 +7373,10 @@ void register_cp_regs_for_features(ARMCPU *cpu) define_arm_cp_regs(cpu, sctlr2_reginfo); } + if (cpu_isar_feature(aa64_tcr2, cpu)) { + define_arm_cp_regs(cpu, tcr2_reginfo); + } + if (cpu_isar_feature(any_predinv, cpu)) { define_arm_cp_regs(cpu, predinv_reginfo); } diff --git a/target/arm/internals.h b/target/arm/internals.h index fb72236255..f5a1e75db3 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -196,6 +196,24 @@ FIELD(CPTR_EL3, TCPAC, 31, 1) #define TTBCR_SH1 (1U << 28) #define TTBCR_EAE (1U << 31) +#define TCR2_PNCH (1ULL << 0) +#define TCR2_PIE (1ULL << 1) +#define TCR2_E0POE (1ULL << 2) +#define TCR2_POE (1ULL << 3) +#define TCR2_AIE (1ULL << 4) +#define TCR2_D128 (1ULL << 5) +#define TCR2_PTTWI (1ULL << 10) +#define TCR2_HAFT (1ULL << 11) +#define TCR2_AMEC0 (1ULL << 12) +#define TCR2_AMEC1 (1ULL << 13) +#define TCR2_DISCH0 (1ULL << 14) +#define TCR2_DISCH1 (1ULL << 15) +#define TCR2_A2 (1ULL << 16) +#define TCR2_FNG0 (1ULL << 17) +#define TCR2_FNG1 (1ULL << 18) +#define TCR2_FNGNA0 (1ULL << 20) +#define TCR2_FNGNA1 (1ULL << 21) + FIELD(VTCR, T0SZ, 0, 6) FIELD(VTCR, SL0, 6, 2) FIELD(VTCR, IRGN0, 8, 2) @@ -230,6 +248,7 @@ FIELD(VSTCR, SA, 30, 1) #define HCRX_CMOW (1ULL << 9) #define HCRX_MCE2 (1ULL << 10) #define HCRX_MSCEN (1ULL << 11) +#define HCRX_TCR2EN (1ULL << 14) #define HCRX_SCTLR2EN (1ULL << 15) #define HPFAR_NS (1ULL << 63) diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c index f4efff03a5..4eb51420ef 100644 --- a/target/arm/tcg/cpu64.c +++ b/target/arm/tcg/cpu64.c @@ -1248,6 +1248,7 @@ void aarch64_max_tcg_initfn(Object *obj) SET_IDREG(isar, ID_AA64MMFR2, t); t = GET_IDREG(isar, ID_AA64MMFR3); + t = FIELD_DP64(t, ID_AA64MMFR3, TCRX, 1); /* FEAT_TCR2 */ t = FIELD_DP64(t, ID_AA64MMFR3, SCTLRX, 1); /* FEAT_SCTLR2 */ t = FIELD_DP64(t, ID_AA64MMFR3, SPEC_FPACC, 1); /* FEAT_FPACC_SPEC */ SET_IDREG(isar, ID_AA64MMFR3, t); From 376cdd7e9c94f1e03b2c58e068e8ebfe78b49514 Mon Sep 17 00:00:00 2001 From: Steve Sistare Date: Tue, 26 Aug 2025 11:21:28 +0100 Subject: [PATCH 06/32] hw/intc/arm_gicv3_kvm: preserve pending interrupts during cpr Close a race condition that causes cpr-transfer to lose VFIO interrupts on ARM. CPR stops VCPUs but does not disable VFIO interrupts, which may continue to arrive throughout the transition to new QEMU. CPR calls kvm_irqchip_remove_irqfd_notifier_gsi in old QEMU to force future interrupts to the producer eventfd, where they are preserved. Old QEMU then destroys the old KVM instance. However, interrupts may already be pending in KVM state. To preserve them, call ioctl KVM_DEV_ARM_VGIC_SAVE_PENDING_TABLES to flush them to guest RAM, where they will be picked up when the new KVM+VCPU instance is created. Cc: qemu-stable@nongnu.org Signed-off-by: Steve Sistare Reviewed-by: Fabiano Rosas Message-id: 1754936384-278328-1-git-send-email-steven.sistare@oracle.com Reviewed-by: Peter Maydell Signed-off-by: Peter Maydell --- hw/intc/arm_gicv3_kvm.c | 15 +++++++++++++++ include/hw/intc/arm_gicv3_common.h | 3 +++ 2 files changed, 18 insertions(+) diff --git a/hw/intc/arm_gicv3_kvm.c b/hw/intc/arm_gicv3_kvm.c index 6166283cd1..0cd14d78a7 100644 --- a/hw/intc/arm_gicv3_kvm.c +++ b/hw/intc/arm_gicv3_kvm.c @@ -31,6 +31,7 @@ #include "gicv3_internal.h" #include "vgic_common.h" #include "migration/blocker.h" +#include "migration/misc.h" #include "qom/object.h" #include "target/arm/cpregs.h" @@ -776,6 +777,17 @@ static void vm_change_state_handler(void *opaque, bool running, } } +static int kvm_arm_gicv3_notifier(NotifierWithReturn *notifier, + MigrationEvent *e, Error **errp) +{ + if (e->type == MIG_EVENT_PRECOPY_DONE) { + GICv3State *s = container_of(notifier, GICv3State, cpr_notifier); + return kvm_device_access(s->dev_fd, KVM_DEV_ARM_VGIC_GRP_CTRL, + KVM_DEV_ARM_VGIC_SAVE_PENDING_TABLES, + NULL, true, errp); + } + return 0; +} static void kvm_arm_gicv3_realize(DeviceState *dev, Error **errp) { @@ -917,6 +929,9 @@ static void kvm_arm_gicv3_realize(DeviceState *dev, Error **errp) if (kvm_device_check_attr(s->dev_fd, KVM_DEV_ARM_VGIC_GRP_CTRL, KVM_DEV_ARM_VGIC_SAVE_PENDING_TABLES)) { qemu_add_vm_change_state_handler(vm_change_state_handler, s); + migration_add_notifier_mode(&s->cpr_notifier, + kvm_arm_gicv3_notifier, + MIG_MODE_CPR_TRANSFER); } } diff --git a/include/hw/intc/arm_gicv3_common.h b/include/hw/intc/arm_gicv3_common.h index c18503869f..572d971d22 100644 --- a/include/hw/intc/arm_gicv3_common.h +++ b/include/hw/intc/arm_gicv3_common.h @@ -27,6 +27,7 @@ #include "hw/sysbus.h" #include "hw/intc/arm_gic_common.h" #include "qom/object.h" +#include "qemu/notify.h" /* * Maximum number of possible interrupts, determined by the GIC architecture. @@ -271,6 +272,8 @@ struct GICv3State { GICv3CPUState *cpu; /* List of all ITSes connected to this GIC */ GPtrArray *itslist; + + NotifierWithReturn cpr_notifier; }; #define GICV3_BITMAP_ACCESSORS(BMP) \ From 186db6a73bc5c01026bb9f4f4a59e442c0156841 Mon Sep 17 00:00:00 2001 From: Smail AIDER Date: Tue, 26 Aug 2025 11:21:28 +0100 Subject: [PATCH 07/32] target/arm: Trap PMCR when MDCR_EL2.TPMCR is set Trap PMCR_EL0 or PMCR accesses to EL2 when MDCR_EL2.TPMCR is set. Similar to MDCR_EL2.TPM, MDCR_EL2.TPMCR allows trapping EL0 and EL1 accesses to the PMCR register to EL2. Cc: qemu-stable@nongnu.org Signed-off-by: Smail AIDER Reviewed-by: Richard Henderson Message-id: 20250811112143.1577055-2-smail.aider@huawei.com Message-Id: <20250722131925.2119169-1-smail.aider@huawei.com> Signed-off-by: Peter Maydell --- target/arm/cpregs-pmu.c | 34 ++++++++++++++++++++++++++-------- 1 file changed, 26 insertions(+), 8 deletions(-) diff --git a/target/arm/cpregs-pmu.c b/target/arm/cpregs-pmu.c index 9c4431c18b..31c01eddc8 100644 --- a/target/arm/cpregs-pmu.c +++ b/target/arm/cpregs-pmu.c @@ -228,22 +228,27 @@ static bool event_supported(uint16_t number) return supported_event_map[number] != UNSUPPORTED_EVENT; } -static CPAccessResult pmreg_access(CPUARMState *env, const ARMCPRegInfo *ri, - bool isread) +static CPAccessResult do_pmreg_access(CPUARMState *env, bool is_pmcr) { /* * Performance monitor registers user accessibility is controlled - * by PMUSERENR. MDCR_EL2.TPM and MDCR_EL3.TPM allow configurable + * by PMUSERENR. MDCR_EL2.TPM/TPMCR and MDCR_EL3.TPM allow configurable * trapping to EL2 or EL3 for other accesses. */ int el = arm_current_el(env); - uint64_t mdcr_el2 = arm_mdcr_el2_eff(env); if (el == 0 && !(env->cp15.c9_pmuserenr & 1)) { return CP_ACCESS_TRAP_EL1; } - if (el < 2 && (mdcr_el2 & MDCR_TPM)) { - return CP_ACCESS_TRAP_EL2; + if (el < 2) { + uint64_t mdcr_el2 = arm_mdcr_el2_eff(env); + + if (mdcr_el2 & MDCR_TPM) { + return CP_ACCESS_TRAP_EL2; + } + if (is_pmcr && (mdcr_el2 & MDCR_TPMCR)) { + return CP_ACCESS_TRAP_EL2; + } } if (el < 3 && (env->cp15.mdcr_el3 & MDCR_TPM)) { return CP_ACCESS_TRAP_EL3; @@ -252,6 +257,19 @@ static CPAccessResult pmreg_access(CPUARMState *env, const ARMCPRegInfo *ri, return CP_ACCESS_OK; } +static CPAccessResult pmreg_access(CPUARMState *env, const ARMCPRegInfo *ri, + bool isread) +{ + return do_pmreg_access(env, false); +} + +static CPAccessResult pmreg_access_pmcr(CPUARMState *env, + const ARMCPRegInfo *ri, + bool isread) +{ + return do_pmreg_access(env, true); +} + static CPAccessResult pmreg_access_xevcntr(CPUARMState *env, const ARMCPRegInfo *ri, bool isread) @@ -1187,14 +1205,14 @@ void define_pm_cpregs(ARMCPU *cpu) .fgt = FGT_PMCR_EL0, .type = ARM_CP_IO | ARM_CP_ALIAS, .fieldoffset = offsetoflow32(CPUARMState, cp15.c9_pmcr), - .accessfn = pmreg_access, + .accessfn = pmreg_access_pmcr, .readfn = pmcr_read, .raw_readfn = raw_read, .writefn = pmcr_write, .raw_writefn = raw_write, }; const ARMCPRegInfo pmcr64 = { .name = "PMCR_EL0", .state = ARM_CP_STATE_AA64, .opc0 = 3, .opc1 = 3, .crn = 9, .crm = 12, .opc2 = 0, - .access = PL0_RW, .accessfn = pmreg_access, + .access = PL0_RW, .accessfn = pmreg_access_pmcr, .fgt = FGT_PMCR_EL0, .type = ARM_CP_IO, .fieldoffset = offsetof(CPUARMState, cp15.c9_pmcr), From 3b53af353b0b2e00fced6f00de87c03346542665 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Tue, 26 Aug 2025 11:21:29 +0100 Subject: [PATCH 08/32] target/arm: Add feature predicate for FEAT_CSSC Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/cpu-features.h | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index 8ec8c3feb3..41511d0835 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -604,6 +604,11 @@ static inline bool isar_feature_aa64_rpres(const ARMISARegisters *id) return FIELD_EX64_IDREG(id, ID_AA64ISAR2, RPRES); } +static inline bool isar_feature_aa64_cssc(const ARMISARegisters *id) +{ + return FIELD_EX64_IDREG(id, ID_AA64ISAR2, CSSC) != 0; +} + static inline bool isar_feature_aa64_lut(const ARMISARegisters *id) { return FIELD_EX64_IDREG(id, ID_AA64ISAR2, LUT); From d0e4b9d4d77ab3685fc22b71de0f4fd220afa17a Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Tue, 26 Aug 2025 11:21:29 +0100 Subject: [PATCH 09/32] target/arm: Implement MIN/MAX (immediate) Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/tcg/a64.decode | 10 ++++++++ target/arm/tcg/translate-a64.c | 44 ++++++++++++++++++++++++++++++++++ 2 files changed, 54 insertions(+) diff --git a/target/arm/tcg/a64.decode b/target/arm/tcg/a64.decode index 8c798cde2b..c1811b0274 100644 --- a/target/arm/tcg/a64.decode +++ b/target/arm/tcg/a64.decode @@ -156,6 +156,16 @@ MOVZ . 10 100101 .. ................ ..... @movw_32 MOVK . 11 100101 .. ................ ..... @movw_64 MOVK . 11 100101 .. ................ ..... @movw_32 +# Min/Max (immediate) + +@minmaxi_s sf:1 .. ........... imm:s8 rn:5 rd:5 &rri_sf +@minmaxi_u sf:1 .. ........... imm:8 rn:5 rd:5 &rri_sf + +SMAX_i . 00 1000111 0000 ........ ..... ..... @minmaxi_s +SMIN_i . 00 1000111 0010 ........ ..... ..... @minmaxi_s +UMAX_i . 00 1000111 0001 ........ ..... ..... @minmaxi_u +UMIN_i . 00 1000111 0011 ........ ..... ..... @minmaxi_u + # Bitfield &bitfield rd rn sf immr imms diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index dbf47595db..b70ae5befd 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -4552,6 +4552,50 @@ TRANS(SUB_i, gen_rri, a, 1, 1, tcg_gen_sub_i64) TRANS(ADDS_i, gen_rri, a, 0, 1, a->sf ? gen_add64_CC : gen_add32_CC) TRANS(SUBS_i, gen_rri, a, 0, 1, a->sf ? gen_sub64_CC : gen_sub32_CC) +/* + * Min/Max (immediate) + */ + +static void gen_wrap3_i32(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, NeonGenTwoOpFn fn) +{ + TCGv_i32 t1 = tcg_temp_new_i32(); + TCGv_i32 t2 = tcg_temp_new_i32(); + + tcg_gen_extrl_i64_i32(t1, n); + tcg_gen_extrl_i64_i32(t2, m); + fn(t1, t1, t2); + tcg_gen_extu_i32_i64(d, t1); +} + +static void gen_smax32_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m) +{ + gen_wrap3_i32(d, n, m, tcg_gen_smax_i32); +} + +static void gen_smin32_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m) +{ + gen_wrap3_i32(d, n, m, tcg_gen_smin_i32); +} + +static void gen_umax32_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m) +{ + gen_wrap3_i32(d, n, m, tcg_gen_umax_i32); +} + +static void gen_umin32_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m) +{ + gen_wrap3_i32(d, n, m, tcg_gen_umin_i32); +} + +TRANS_FEAT(SMAX_i, aa64_cssc, gen_rri, a, 0, 0, + a->sf ? tcg_gen_smax_i64 : gen_smax32_i64) +TRANS_FEAT(SMIN_i, aa64_cssc, gen_rri, a, 0, 0, + a->sf ? tcg_gen_smin_i64 : gen_smin32_i64) +TRANS_FEAT(UMAX_i, aa64_cssc, gen_rri, a, 0, 0, + a->sf ? tcg_gen_umax_i64 : gen_umax32_i64) +TRANS_FEAT(UMIN_i, aa64_cssc, gen_rri, a, 0, 0, + a->sf ? tcg_gen_umin_i64 : gen_umin32_i64) + /* * Add/subtract (immediate, with tags) */ From 506538208ddabc9b15b02a7e865aa1b64f9f18e5 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Tue, 26 Aug 2025 11:21:29 +0100 Subject: [PATCH 10/32] target/arm: Implement MIN/MAX (register) Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/tcg/a64.decode | 5 +++++ target/arm/tcg/translate-a64.c | 22 ++++++++++++++++++++++ 2 files changed, 27 insertions(+) diff --git a/target/arm/tcg/a64.decode b/target/arm/tcg/a64.decode index c1811b0274..a886b3ba4c 100644 --- a/target/arm/tcg/a64.decode +++ b/target/arm/tcg/a64.decode @@ -708,6 +708,11 @@ GMI 1 00 11010110 ..... 000101 ..... ..... @rrr PACGA 1 00 11010110 ..... 001100 ..... ..... @rrr +SMAX . 00 11010110 ..... 011000 ..... ..... @rrr_sf +SMIN . 00 11010110 ..... 011010 ..... ..... @rrr_sf +UMAX . 00 11010110 ..... 011001 ..... ..... @rrr_sf +UMIN . 00 11010110 ..... 011011 ..... ..... @rrr_sf + # Data Processing (1-source) @rr . .......... ..... ...... rn:5 rd:5 &rr diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index b70ae5befd..bb92bdc296 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -8201,6 +8201,28 @@ static bool trans_PACGA(DisasContext *s, arg_rrr *a) return false; } +static bool gen_rrr(DisasContext *s, arg_rrr_sf *a, ArithTwoOp fn) +{ + TCGv_i64 tcg_rm = cpu_reg(s, a->rm); + TCGv_i64 tcg_rn = cpu_reg(s, a->rn); + TCGv_i64 tcg_rd = cpu_reg(s, a->rd); + + fn(tcg_rd, tcg_rn, tcg_rm); + if (!a->sf) { + tcg_gen_ext32u_i64(tcg_rd, tcg_rd); + } + return true; +} + +TRANS_FEAT(SMAX, aa64_cssc, gen_rrr, a, + a->sf ? tcg_gen_smax_i64 : gen_smax32_i64) +TRANS_FEAT(SMIN, aa64_cssc, gen_rrr, a, + a->sf ? tcg_gen_smin_i64 : gen_smin32_i64) +TRANS_FEAT(UMAX, aa64_cssc, gen_rrr, a, + a->sf ? tcg_gen_umax_i64 : gen_umax32_i64) +TRANS_FEAT(UMIN, aa64_cssc, gen_rrr, a, + a->sf ? tcg_gen_umin_i64 : gen_umin32_i64) + typedef void ArithOneOp(TCGv_i64, TCGv_i64); static bool gen_rr(DisasContext *s, int rd, int rn, ArithOneOp fn) From 8a4bb8b975f80022cdea27757cd8d14f13bf65a9 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Tue, 26 Aug 2025 11:21:29 +0100 Subject: [PATCH 11/32] target/arm: Split out gen_wrap2_i32 helper Wrapper to extract the low 32 bits, perform an operation, and zero-extend back to 64 bits. Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-5-richard.henderson@linaro.org [PMM: fixed wrong output-reg argument in callsites; add comment] Signed-off-by: Peter Maydell --- target/arm/tcg/translate-a64.c | 25 +++++++++++++++---------- 1 file changed, 15 insertions(+), 10 deletions(-) diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index bb92bdc296..c0fa9a44e7 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -8231,13 +8231,22 @@ static bool gen_rr(DisasContext *s, int rd, int rn, ArithOneOp fn) return true; } +/* + * Perform 32-bit operation fn on the low half of n; + * the high half of the output is zeroed. + */ +static void gen_wrap2_i32(TCGv_i64 d, TCGv_i64 n, NeonGenOneOpFn fn) +{ + TCGv_i32 t = tcg_temp_new_i32(); + + tcg_gen_extrl_i64_i32(t, n); + fn(t, t); + tcg_gen_extu_i32_i64(d, t); +} + static void gen_rbit32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) { - TCGv_i32 t32 = tcg_temp_new_i32(); - - tcg_gen_extrl_i64_i32(t32, tcg_rn); - gen_helper_rbit(t32, t32); - tcg_gen_extu_i32_i64(tcg_rd, t32); + gen_wrap2_i32(tcg_rd, tcg_rn, gen_helper_rbit); } static void gen_rev16_xx(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn, TCGv_i64 mask) @@ -8293,11 +8302,7 @@ static void gen_clz64(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) static void gen_cls32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) { - TCGv_i32 t32 = tcg_temp_new_i32(); - - tcg_gen_extrl_i64_i32(t32, tcg_rn); - tcg_gen_clrsb_i32(t32, t32); - tcg_gen_extu_i32_i64(tcg_rd, t32); + gen_wrap2_i32(tcg_rd, tcg_rn, tcg_gen_clrsb_i32); } TRANS(CLZ, gen_rr, a->rd, a->rn, a->sf ? gen_clz64 : gen_clz32) From 994a260feac452ca478af5bd4ba4bff45b889b6e Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Tue, 26 Aug 2025 17:11:12 +0100 Subject: [PATCH 12/32] target/arm: Implement CTZ, CNT, ABS Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-6-richard.henderson@linaro.org [PMM: fix tcg_rd/tcg_rn mixup] Signed-off-by: Peter Maydell --- target/arm/tcg/a64.decode | 4 ++++ target/arm/tcg/translate-a64.c | 31 +++++++++++++++++++++++++++++++ 2 files changed, 35 insertions(+) diff --git a/target/arm/tcg/a64.decode b/target/arm/tcg/a64.decode index a886b3ba4c..766c610c01 100644 --- a/target/arm/tcg/a64.decode +++ b/target/arm/tcg/a64.decode @@ -726,6 +726,10 @@ REV64 1 10 11010110 00000 000011 ..... ..... @rr CLZ . 10 11010110 00000 000100 ..... ..... @rr_sf CLS . 10 11010110 00000 000101 ..... ..... @rr_sf +CTZ . 10 11010110 00000 000110 ..... ..... @rr_sf +CNT . 10 11010110 00000 000111 ..... ..... @rr_sf +ABS . 10 11010110 00000 001000 ..... ..... @rr_sf + &pacaut rd rn z @pacaut . .. ........ ..... .. z:1 ... rn:5 rd:5 &pacaut diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index c0fa9a44e7..259aa70a36 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -8308,6 +8308,37 @@ static void gen_cls32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) TRANS(CLZ, gen_rr, a->rd, a->rn, a->sf ? gen_clz64 : gen_clz32) TRANS(CLS, gen_rr, a->rd, a->rn, a->sf ? tcg_gen_clrsb_i64 : gen_cls32) +static void gen_ctz32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) +{ + TCGv_i32 t32 = tcg_temp_new_i32(); + + tcg_gen_extrl_i64_i32(t32, tcg_rn); + tcg_gen_ctzi_i32(t32, t32, 32); + tcg_gen_extu_i32_i64(tcg_rd, t32); +} + +static void gen_ctz64(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) +{ + tcg_gen_ctzi_i64(tcg_rd, tcg_rn, 64); +} + +static void gen_cnt32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) +{ + gen_wrap2_i32(tcg_rd, tcg_rn, tcg_gen_ctpop_i32); +} + +static void gen_abs32(TCGv_i64 tcg_rd, TCGv_i64 tcg_rn) +{ + gen_wrap2_i32(tcg_rd, tcg_rn, tcg_gen_abs_i32); +} + +TRANS_FEAT(CTZ, aa64_cssc, gen_rr, a->rd, a->rn, + a->sf ? gen_ctz64 : gen_ctz32) +TRANS_FEAT(CNT, aa64_cssc, gen_rr, a->rd, a->rn, + a->sf ? tcg_gen_ctpop_i64 : gen_cnt32) +TRANS_FEAT(ABS, aa64_cssc, gen_rr, a->rd, a->rn, + a->sf ? tcg_gen_abs_i64 : gen_abs32) + static bool gen_pacaut(DisasContext *s, arg_pacaut *a, NeonGenTwo64OpEnvFn fn) { TCGv_i64 tcg_rd, tcg_rn; From 7494f8bbfbb030c5c40a42e4d71430115e4f7a63 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Tue, 26 Aug 2025 11:21:30 +0100 Subject: [PATCH 13/32] target/arm: Enable FEAT_CSSC for -cpu max Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250803014019.416797-7-richard.henderson@linaro.org [PMM: rebased to handle linux-user elfload.c refactor] Signed-off-by: Peter Maydell --- docs/system/arm/emulation.rst | 1 + linux-user/aarch64/elfload.c | 1 + target/arm/tcg/cpu64.c | 1 + 3 files changed, 3 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index 1c597d8673..b12f013b4f 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -30,6 +30,7 @@ the following architecture extensions: - FEAT_CMOW (Control for cache maintenance permission) - FEAT_CRC32 (CRC32 instructions) - FEAT_Crypto (Cryptographic Extension) +- FEAT_CSSC (Common Short Sequence Compression instructions) - FEAT_CSV2 (Cache speculation variant 2) - FEAT_CSV2_1p1 (Cache speculation variant 2, version 1.1) - FEAT_CSV2_1p2 (Cache speculation variant 2, version 1.2) diff --git a/linux-user/aarch64/elfload.c b/linux-user/aarch64/elfload.c index 8076968251..dd5f34398a 100644 --- a/linux-user/aarch64/elfload.c +++ b/linux-user/aarch64/elfload.c @@ -215,6 +215,7 @@ abi_ulong get_elf_hwcap2(CPUState *cs) GET_FEATURE_ID(aa64_sme_b16b16, ARM_HWCAP2_A64_SME_B16B16); GET_FEATURE_ID(aa64_sme_f16f16, ARM_HWCAP2_A64_SME_F16F16); GET_FEATURE_ID(aa64_sve_b16b16, ARM_HWCAP2_A64_SVE_B16B16); + GET_FEATURE_ID(aa64_cssc, ARM_HWCAP2_A64_CSSC); return hwcaps; } diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c index 4eb51420ef..eaf8846a6a 100644 --- a/target/arm/tcg/cpu64.c +++ b/target/arm/tcg/cpu64.c @@ -1178,6 +1178,7 @@ void aarch64_max_tcg_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64ISAR2, MOPS, 1); /* FEAT_MOPS */ t = FIELD_DP64(t, ID_AA64ISAR2, BC, 1); /* FEAT_HBC */ t = FIELD_DP64(t, ID_AA64ISAR2, WFXT, 2); /* FEAT_WFxT */ + t = FIELD_DP64(t, ID_AA64ISAR2, CSSC, 1); /* FEAT_CSSC */ SET_IDREG(isar, ID_AA64ISAR2, t); t = GET_IDREG(isar, ID_AA64PFR0); From 36bc78aca83cfd3c8f73cbcb428bc7f0ce7a4a61 Mon Sep 17 00:00:00 2001 From: Manos Pitsidianakis Date: Tue, 26 Aug 2025 11:21:30 +0100 Subject: [PATCH 14/32] hw/arm: add static NVDIMMs in device tree NVDIMM is used for fast rootfs with EROFS, for example by kata containers. To allow booting with static NVDIMM memory, add them to the device tree in arm virt machine. This allows users to boot directly with nvdimm memory devices without having to rely on ACPI and hotplug. Verified to work with command invocation: ./qemu-system-aarch64 \ -M virt,nvdimm=on \ -cpu cortex-a57 \ -m 4G,slots=2,maxmem=8G \ -object memory-backend-file,id=mem1,share=on,mem-path=/tmp/nvdimm,size=4G,readonly=off \ -device nvdimm,id=nvdimm1,memdev=mem1,unarmed=off \ -drive file=./debian-12-nocloud-arm64-commited.qcow2,format=qcow2 \ -kernel ./vmlinuz-6.1.0-13-arm64 \ -append "root=/dev/vda1 console=ttyAMA0,115200 acpi=off" -initrd ./initrd.img-6.1.0-13-arm64 \ -nographic \ -serial mon:stdio Signed-off-by: Manos Pitsidianakis Message-id: 20250807-nvdimm_arm64_virt-v2-1-b8054578bea8@linaro.org Reviewed-by: Peter Maydell Signed-off-by: Peter Maydell --- hw/arm/boot.c | 42 ++++++++++++++++++++++++++++++++++++++++++ hw/arm/virt.c | 8 +++++--- 2 files changed, 47 insertions(+), 3 deletions(-) diff --git a/hw/arm/boot.c b/hw/arm/boot.c index d391cd01bb..1e57c4ab9e 100644 --- a/hw/arm/boot.c +++ b/hw/arm/boot.c @@ -25,6 +25,7 @@ #include "hw/boards.h" #include "system/reset.h" #include "hw/loader.h" +#include "hw/mem/memory-device.h" #include "elf.h" #include "system/device_tree.h" #include "qemu/config-file.h" @@ -515,6 +516,29 @@ static void fdt_add_psci_node(void *fdt, ARMCPU *armcpu) qemu_fdt_setprop_cell(fdt, "/psci", "migrate", migrate_fn); } +static int fdt_add_pmem_node(void *fdt, uint32_t acells, uint32_t scells, + int64_t mem_base, int64_t size, int64_t node) +{ + int ret; + + g_autofree char *nodename = g_strdup_printf("/pmem@%" PRIx64, mem_base); + + qemu_fdt_add_subnode(fdt, nodename); + qemu_fdt_setprop_string(fdt, nodename, "compatible", "pmem-region"); + ret = qemu_fdt_setprop_sized_cells(fdt, nodename, "reg", acells, + mem_base, scells, size); + if (ret) { + return ret; + } + + if (node >= 0) { + return qemu_fdt_setprop_cell(fdt, nodename, "numa-node-id", + node); + } + + return 0; +} + int arm_load_dtb(hwaddr addr, const struct arm_boot_info *binfo, hwaddr addr_limit, AddressSpace *as, MachineState *ms, ARMCPU *cpu) @@ -525,6 +549,7 @@ int arm_load_dtb(hwaddr addr, const struct arm_boot_info *binfo, unsigned int i; hwaddr mem_base, mem_len; char **node_path; + g_autofree MemoryDeviceInfoList *md_list = NULL; Error *err = NULL; if (binfo->dtb_filename) { @@ -628,6 +653,23 @@ int arm_load_dtb(hwaddr addr, const struct arm_boot_info *binfo, } } + md_list = qmp_memory_device_list(); + for (MemoryDeviceInfoList *m = md_list; m != NULL; m = m->next) { + MemoryDeviceInfo *mi = m->value; + + if (mi->type == MEMORY_DEVICE_INFO_KIND_NVDIMM) { + PCDIMMDeviceInfo *di = mi->u.nvdimm.data; + + rc = fdt_add_pmem_node(fdt, acells, scells, + di->addr, di->size, di->node); + if (rc < 0) { + fprintf(stderr, "couldn't add NVDIMM /pmem@%"PRIx64" node\n", + di->addr); + goto fail; + } + } + } + rc = fdt_path_offset(fdt, "/chosen"); if (rc < 0) { qemu_fdt_add_subnode(fdt, "/chosen"); diff --git a/hw/arm/virt.c b/hw/arm/virt.c index 9326cfc895..1e63f40fbe 100644 --- a/hw/arm/virt.c +++ b/hw/arm/virt.c @@ -2917,7 +2917,7 @@ static void virt_memory_pre_plug(HotplugHandler *hotplug_dev, DeviceState *dev, const MachineState *ms = MACHINE(hotplug_dev); const bool is_nvdimm = object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM); - if (!vms->acpi_dev) { + if (!vms->acpi_dev && !(is_nvdimm && !dev->hotplugged)) { error_setg(errp, "memory hotplug is not enabled: missing acpi-ged device"); return; @@ -2949,8 +2949,10 @@ static void virt_memory_plug(HotplugHandler *hotplug_dev, nvdimm_plug(ms->nvdimms_state); } - hotplug_handler_plug(HOTPLUG_HANDLER(vms->acpi_dev), - dev, &error_abort); + if (vms->acpi_dev) { + hotplug_handler_plug(HOTPLUG_HANDLER(vms->acpi_dev), + dev, &error_abort); + } } static void virt_machine_device_pre_plug_cb(HotplugHandler *hotplug_dev, From 5ffd387e9e0f787744fadaad35e1bf92224b0642 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Tue, 19 Aug 2025 12:56:48 +0100 Subject: [PATCH 15/32] scripts/kernel-doc: Avoid new Perl precedence warning MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Newer versions of Perl (5.41.x and up) emit a warning for code in kernel-doc: Possible precedence problem between ! and pattern match (m//) at /scripts/kernel-doc line 1597. This is because the code does: if (!$param =~ /\w\.\.\.$/) { In Perl, the ! operator has higher precedence than the =~ pattern-match binding, so the effect of this condition is to first logically-negate the string $param into a true-or-false value and then try to pattern match it against the regex, which in this case will always fail. This is almost certainly not what the author intended. In the new Python version of kernel-doc in the Linux kernel, the equivalent code is written: if KernRe(r'\w\.\.\.$').search(param): # For named variable parameters of the form `x...`, # remove the dots param = param[:-3] else: # Handles unnamed variable parameters param = "..." which is a more sensible way of writing the behaviour you would get if you put in brackets to make the regex match first and then negate the result. Take this as the intended behaviour, and update the Perl to match. For QEMU, this produces no change in output, presumably because we never used the "unnamed variable parameters" syntax. Cc: qemu-stable@nongnu.org Signed-off-by: Peter Maydell Reviewed-by: Daniel P. Berrangé Reviewed-by: Mauro Carvalho Chehab Message-id: 20250819115648.2125709-1-peter.maydell@linaro.org --- scripts/kernel-doc | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/scripts/kernel-doc b/scripts/kernel-doc index fec83f53ed..117ec8fcd1 100755 --- a/scripts/kernel-doc +++ b/scripts/kernel-doc @@ -1594,13 +1594,12 @@ sub push_parameter($$$$$) { if ($type eq "" && $param =~ /\.\.\.$/) { - if (!$param =~ /\w\.\.\.$/) { - # handles unnamed variable parameters - $param = "..."; - } - elsif ($param =~ /\w\.\.\.$/) { + if ($param =~ /\w\.\.\.$/) { # for named variable parameters of the form `x...`, remove the dots $param =~ s/\.\.\.$//; + } else { + # handles unnamed variable parameters + $param = "..."; } if (!defined $parameterdescs{$param} || $parameterdescs{$param} eq "") { $parameterdescs{$param} = "variable arguments"; From 3f34478007a2423399dbdfae5e3deeca3898da9c Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:16 +0100 Subject: [PATCH 16/32] docs/sphinx/kerneldoc.py: Handle new LINENO syntax The new upstream kernel-doc that we plan to update to uses a different syntax for the LINENO directives that the Sphinx extension parses: instead of #define LINENO 86 it has .. LINENO 86 Update the kerneldoc.py extension to handle both syntaxes, so that it will work with both the old and the new kernel-doc. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-2-peter.maydell@linaro.org --- docs/sphinx/kerneldoc.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/docs/sphinx/kerneldoc.py b/docs/sphinx/kerneldoc.py index 3aa972f2e8..30bb343198 100644 --- a/docs/sphinx/kerneldoc.py +++ b/docs/sphinx/kerneldoc.py @@ -127,7 +127,7 @@ class KernelDocDirective(Directive): result = ViewList() lineoffset = 0; - line_regex = re.compile("^#define LINENO ([0-9]+)$") + line_regex = re.compile(r"^(?:\.\.|#define) LINENO ([0-9]+)$") for line in lines: match = line_regex.search(line) if match: From 4a1bc66d3f61cb803b16bb1691e46b5ec59d25f5 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:17 +0100 Subject: [PATCH 17/32] tests/qtest/libqtest.h: Remove stray space from doc comment The doc comment for qtest_cb_for_every_machine has a stray space at the start of its description, which makes kernel-doc think that this line is part of the documentation of the skip_old_versioned argument. The result is that the HTML doesn't have a "Description" section and the text is instead put in the wrong place. Remove the stray space. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-3-peter.maydell@linaro.org --- tests/qtest/libqtest.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/qtest/libqtest.h b/tests/qtest/libqtest.h index b3f2e7fbef..fd27521a9c 100644 --- a/tests/qtest/libqtest.h +++ b/tests/qtest/libqtest.h @@ -977,7 +977,7 @@ void qtest_qmp_fds_assert_success(QTestState *qts, int *fds, size_t nfds, * @cb: Pointer to the callback function * @skip_old_versioned: true if versioned old machine types should be skipped * - * Call a callback function for every name of all available machines. + * Call a callback function for every name of all available machines. */ void qtest_cb_for_every_machine(void (*cb)(const char *machine), bool skip_old_versioned); From 2b2765ac4045642563cc92ad98c2244a0aa0c7fc Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:18 +0100 Subject: [PATCH 18/32] scripts: Import Python kerneldoc from Linux kernel We last synced our copy of kerneldoc with Linux back in 2020. In the interim, upstream has entirely rewritten the script in Python, and the new Python version is split into a main script plus some libraries in the kernel's scripts/lib/kdoc. Import all these files. These are the versions as of kernel commit 0cc53520e68be, with no local changes. We use the same lib/kdoc/ directory as the kernel does here, so we can avoid having to edit the top-level script just to adjust a pathname, even though it is probably not the naming we would have picked if this was a purely QEMU script. The Sphinx conf.py still points at the Perl version of the script, so this Python code will not be invoked to build the docs yet. NB: checkpatch complains about many things in this commit, including the use of "GPL-2.0" rather than "GPL-2.0-only" in the SPDX tags, but since this is a third party import we can ignore this. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-4-peter.maydell@linaro.org --- scripts/kernel-doc.py | 325 ++++++ scripts/lib/kdoc/kdoc_files.py | 291 ++++++ scripts/lib/kdoc/kdoc_item.py | 42 + scripts/lib/kdoc/kdoc_output.py | 749 ++++++++++++++ scripts/lib/kdoc/kdoc_parser.py | 1669 +++++++++++++++++++++++++++++++ scripts/lib/kdoc/kdoc_re.py | 270 +++++ 6 files changed, 3346 insertions(+) create mode 100755 scripts/kernel-doc.py create mode 100644 scripts/lib/kdoc/kdoc_files.py create mode 100644 scripts/lib/kdoc/kdoc_item.py create mode 100644 scripts/lib/kdoc/kdoc_output.py create mode 100644 scripts/lib/kdoc/kdoc_parser.py create mode 100644 scripts/lib/kdoc/kdoc_re.py diff --git a/scripts/kernel-doc.py b/scripts/kernel-doc.py new file mode 100755 index 0000000000..fc3d46ef51 --- /dev/null +++ b/scripts/kernel-doc.py @@ -0,0 +1,325 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +# Copyright(c) 2025: Mauro Carvalho Chehab . +# +# pylint: disable=C0103,R0915 +# +# Converted from the kernel-doc script originally written in Perl +# under GPLv2, copyrighted since 1998 by the following authors: +# +# Aditya Srivastava +# Akira Yokosawa +# Alexander A. Klimov +# Alexander Lobakin +# André Almeida +# Andy Shevchenko +# Anna-Maria Behnsen +# Armin Kuster +# Bart Van Assche +# Ben Hutchings +# Borislav Petkov +# Chen-Yu Tsai +# Coco Li +# Conchúr Navid +# Daniel Santos +# Danilo Cesar Lemes de Paula +# Dan Luedtke +# Donald Hunter +# Gabriel Krisman Bertazi +# Greg Kroah-Hartman +# Harvey Harrison +# Horia Geanta +# Ilya Dryomov +# Jakub Kicinski +# Jani Nikula +# Jason Baron +# Jason Gunthorpe +# Jérémy Bobbio +# Johannes Berg +# Johannes Weiner +# Jonathan Cameron +# Jonathan Corbet +# Jonathan Neuschäfer +# Kamil Rytarowski +# Kees Cook +# Laurent Pinchart +# Levin, Alexander (Sasha Levin) +# Linus Torvalds +# Lucas De Marchi +# Mark Rutland +# Markus Heiser +# Martin Waitz +# Masahiro Yamada +# Matthew Wilcox +# Mauro Carvalho Chehab +# Michal Wajdeczko +# Michael Zucchi +# Mike Rapoport +# Niklas Söderlund +# Nishanth Menon +# Paolo Bonzini +# Pavan Kumar Linga +# Pavel Pisa +# Peter Maydell +# Pierre-Louis Bossart +# Randy Dunlap +# Richard Kennedy +# Rich Walker +# Rolf Eike Beer +# Sakari Ailus +# Silvio Fricke +# Simon Huggins +# Tim Waugh +# Tomasz Warniełło +# Utkarsh Tripathi +# valdis.kletnieks@vt.edu +# Vegard Nossum +# Will Deacon +# Yacine Belkadi +# Yujie Liu + +""" +kernel_doc +========== + +Print formatted kernel documentation to stdout + +Read C language source or header FILEs, extract embedded +documentation comments, and print formatted documentation +to standard output. + +The documentation comments are identified by the "/**" +opening comment mark. + +See Documentation/doc-guide/kernel-doc.rst for the +documentation comment syntax. +""" + +import argparse +import logging +import os +import sys + +# Import Python modules + +LIB_DIR = "lib/kdoc" +SRC_DIR = os.path.dirname(os.path.realpath(__file__)) + +sys.path.insert(0, os.path.join(SRC_DIR, LIB_DIR)) + +from kdoc_files import KernelFiles # pylint: disable=C0413 +from kdoc_output import RestFormat, ManFormat # pylint: disable=C0413 + +DESC = """ +Read C language source or header FILEs, extract embedded documentation comments, +and print formatted documentation to standard output. + +The documentation comments are identified by the "/**" opening comment mark. + +See Documentation/doc-guide/kernel-doc.rst for the documentation comment syntax. +""" + +EXPORT_FILE_DESC = """ +Specify an additional FILE in which to look for EXPORT_SYMBOL information. + +May be used multiple times. +""" + +EXPORT_DESC = """ +Only output documentation for the symbols that have been +exported using EXPORT_SYMBOL() and related macros in any input +FILE or -export-file FILE. +""" + +INTERNAL_DESC = """ +Only output documentation for the symbols that have NOT been +exported using EXPORT_SYMBOL() and related macros in any input +FILE or -export-file FILE. +""" + +FUNCTION_DESC = """ +Only output documentation for the given function or DOC: section +title. All other functions and DOC: sections are ignored. + +May be used multiple times. +""" + +NOSYMBOL_DESC = """ +Exclude the specified symbol from the output documentation. + +May be used multiple times. +""" + +FILES_DESC = """ +Header and C source files to be parsed. +""" + +WARN_CONTENTS_BEFORE_SECTIONS_DESC = """ +Warns if there are contents before sections (deprecated). + +This option is kept just for backward-compatibility, but it does nothing, +neither here nor at the original Perl script. +""" + + +class MsgFormatter(logging.Formatter): + """Helper class to format warnings on a similar way to kernel-doc.pl""" + + def format(self, record): + record.levelname = record.levelname.capitalize() + return logging.Formatter.format(self, record) + +def main(): + """Main program""" + + parser = argparse.ArgumentParser(formatter_class=argparse.RawTextHelpFormatter, + description=DESC) + + # Normal arguments + + parser.add_argument("-v", "-verbose", "--verbose", action="store_true", + help="Verbose output, more warnings and other information.") + + parser.add_argument("-d", "-debug", "--debug", action="store_true", + help="Enable debug messages") + + parser.add_argument("-M", "-modulename", "--modulename", + default="Kernel API", + help="Allow setting a module name at the output.") + + parser.add_argument("-l", "-enable-lineno", "--enable_lineno", + action="store_true", + help="Enable line number output (only in ReST mode)") + + # Arguments to control the warning behavior + + parser.add_argument("-Wreturn", "--wreturn", action="store_true", + help="Warns about the lack of a return markup on functions.") + + parser.add_argument("-Wshort-desc", "-Wshort-description", "--wshort-desc", + action="store_true", + help="Warns if initial short description is missing") + + parser.add_argument("-Wcontents-before-sections", + "--wcontents-before-sections", action="store_true", + help=WARN_CONTENTS_BEFORE_SECTIONS_DESC) + + parser.add_argument("-Wall", "--wall", action="store_true", + help="Enable all types of warnings") + + parser.add_argument("-Werror", "--werror", action="store_true", + help="Treat warnings as errors.") + + parser.add_argument("-export-file", "--export-file", action='append', + help=EXPORT_FILE_DESC) + + # Output format mutually-exclusive group + + out_group = parser.add_argument_group("Output format selection (mutually exclusive)") + + out_fmt = out_group.add_mutually_exclusive_group() + + out_fmt.add_argument("-m", "-man", "--man", action="store_true", + help="Output troff manual page format.") + out_fmt.add_argument("-r", "-rst", "--rst", action="store_true", + help="Output reStructuredText format (default).") + out_fmt.add_argument("-N", "-none", "--none", action="store_true", + help="Do not output documentation, only warnings.") + + # Output selection mutually-exclusive group + + sel_group = parser.add_argument_group("Output selection (mutually exclusive)") + sel_mut = sel_group.add_mutually_exclusive_group() + + sel_mut.add_argument("-e", "-export", "--export", action='store_true', + help=EXPORT_DESC) + + sel_mut.add_argument("-i", "-internal", "--internal", action='store_true', + help=INTERNAL_DESC) + + sel_mut.add_argument("-s", "-function", "--symbol", action='append', + help=FUNCTION_DESC) + + # Those are valid for all 3 types of filter + parser.add_argument("-n", "-nosymbol", "--nosymbol", action='append', + help=NOSYMBOL_DESC) + + parser.add_argument("-D", "-no-doc-sections", "--no-doc-sections", + action='store_true', help="Don't outputt DOC sections") + + parser.add_argument("files", metavar="FILE", + nargs="+", help=FILES_DESC) + + args = parser.parse_args() + + if args.wall: + args.wreturn = True + args.wshort_desc = True + args.wcontents_before_sections = True + + logger = logging.getLogger() + + if not args.debug: + logger.setLevel(logging.INFO) + else: + logger.setLevel(logging.DEBUG) + + formatter = MsgFormatter('%(levelname)s: %(message)s') + + handler = logging.StreamHandler() + handler.setFormatter(formatter) + + logger.addHandler(handler) + + python_ver = sys.version_info[:2] + if python_ver < (3,6): + logger.warning("Python 3.6 or later is required by kernel-doc") + + # Return 0 here to avoid breaking compilation + sys.exit(0) + + if python_ver < (3,7): + logger.warning("Python 3.7 or later is required for correct results") + + if args.man: + out_style = ManFormat(modulename=args.modulename) + elif args.none: + out_style = None + else: + out_style = RestFormat() + + kfiles = KernelFiles(verbose=args.verbose, + out_style=out_style, werror=args.werror, + wreturn=args.wreturn, wshort_desc=args.wshort_desc, + wcontents_before_sections=args.wcontents_before_sections) + + kfiles.parse(args.files, export_file=args.export_file) + + for t in kfiles.msg(enable_lineno=args.enable_lineno, export=args.export, + internal=args.internal, symbol=args.symbol, + nosymbol=args.nosymbol, export_file=args.export_file, + no_doc_sections=args.no_doc_sections): + msg = t[1] + if msg: + print(msg) + + error_count = kfiles.errors + if not error_count: + sys.exit(0) + + if args.werror: + print(f"{error_count} warnings as errors") + sys.exit(error_count) + + if args.verbose: + print(f"{error_count} errors") + + if args.none: + sys.exit(0) + + sys.exit(error_count) + + +# Call main method +if __name__ == "__main__": + main() diff --git a/scripts/lib/kdoc/kdoc_files.py b/scripts/lib/kdoc/kdoc_files.py new file mode 100644 index 0000000000..9e09b45b02 --- /dev/null +++ b/scripts/lib/kdoc/kdoc_files.py @@ -0,0 +1,291 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +# Copyright(c) 2025: Mauro Carvalho Chehab . +# +# pylint: disable=R0903,R0913,R0914,R0917 + +""" +Parse lernel-doc tags on multiple kernel source files. +""" + +import argparse +import logging +import os +import re + +from kdoc_parser import KernelDoc +from kdoc_output import OutputFormat + + +class GlobSourceFiles: + """ + Parse C source code file names and directories via an Interactor. + """ + + def __init__(self, srctree=None, valid_extensions=None): + """ + Initialize valid extensions with a tuple. + + If not defined, assume default C extensions (.c and .h) + + It would be possible to use python's glob function, but it is + very slow, and it is not interactive. So, it would wait to read all + directories before actually do something. + + So, let's use our own implementation. + """ + + if not valid_extensions: + self.extensions = (".c", ".h") + else: + self.extensions = valid_extensions + + self.srctree = srctree + + def _parse_dir(self, dirname): + """Internal function to parse files recursively""" + + with os.scandir(dirname) as obj: + for entry in obj: + name = os.path.join(dirname, entry.name) + + if entry.is_dir(): + yield from self._parse_dir(name) + + if not entry.is_file(): + continue + + basename = os.path.basename(name) + + if not basename.endswith(self.extensions): + continue + + yield name + + def parse_files(self, file_list, file_not_found_cb): + """ + Define an interator to parse all source files from file_list, + handling directories if any + """ + + if not file_list: + return + + for fname in file_list: + if self.srctree: + f = os.path.join(self.srctree, fname) + else: + f = fname + + if os.path.isdir(f): + yield from self._parse_dir(f) + elif os.path.isfile(f): + yield f + elif file_not_found_cb: + file_not_found_cb(fname) + + +class KernelFiles(): + """ + Parse kernel-doc tags on multiple kernel source files. + + There are two type of parsers defined here: + - self.parse_file(): parses both kernel-doc markups and + EXPORT_SYMBOL* macros; + - self.process_export_file(): parses only EXPORT_SYMBOL* macros. + """ + + def warning(self, msg): + """Ancillary routine to output a warning and increment error count""" + + self.config.log.warning(msg) + self.errors += 1 + + def error(self, msg): + """Ancillary routine to output an error and increment error count""" + + self.config.log.error(msg) + self.errors += 1 + + def parse_file(self, fname): + """ + Parse a single Kernel source. + """ + + # Prevent parsing the same file twice if results are cached + if fname in self.files: + return + + doc = KernelDoc(self.config, fname) + export_table, entries = doc.parse_kdoc() + + self.export_table[fname] = export_table + + self.files.add(fname) + self.export_files.add(fname) # parse_kdoc() already check exports + + self.results[fname] = entries + + def process_export_file(self, fname): + """ + Parses EXPORT_SYMBOL* macros from a single Kernel source file. + """ + + # Prevent parsing the same file twice if results are cached + if fname in self.export_files: + return + + doc = KernelDoc(self.config, fname) + export_table = doc.parse_export() + + if not export_table: + self.error(f"Error: Cannot check EXPORT_SYMBOL* on {fname}") + export_table = set() + + self.export_table[fname] = export_table + self.export_files.add(fname) + + def file_not_found_cb(self, fname): + """ + Callback to warn if a file was not found. + """ + + self.error(f"Cannot find file {fname}") + + def __init__(self, verbose=False, out_style=None, + werror=False, wreturn=False, wshort_desc=False, + wcontents_before_sections=False, + logger=None): + """ + Initialize startup variables and parse all files + """ + + if not verbose: + verbose = bool(os.environ.get("KBUILD_VERBOSE", 0)) + + if out_style is None: + out_style = OutputFormat() + + if not werror: + kcflags = os.environ.get("KCFLAGS", None) + if kcflags: + match = re.search(r"(\s|^)-Werror(\s|$)/", kcflags) + if match: + werror = True + + # reading this variable is for backwards compat just in case + # someone was calling it with the variable from outside the + # kernel's build system + kdoc_werror = os.environ.get("KDOC_WERROR", None) + if kdoc_werror: + werror = kdoc_werror + + # Some variables are global to the parser logic as a whole as they are + # used to send control configuration to KernelDoc class. As such, + # those variables are read-only inside the KernelDoc. + self.config = argparse.Namespace + + self.config.verbose = verbose + self.config.werror = werror + self.config.wreturn = wreturn + self.config.wshort_desc = wshort_desc + self.config.wcontents_before_sections = wcontents_before_sections + + if not logger: + self.config.log = logging.getLogger("kernel-doc") + else: + self.config.log = logger + + self.config.warning = self.warning + + self.config.src_tree = os.environ.get("SRCTREE", None) + + # Initialize variables that are internal to KernelFiles + + self.out_style = out_style + + self.errors = 0 + self.results = {} + + self.files = set() + self.export_files = set() + self.export_table = {} + + def parse(self, file_list, export_file=None): + """ + Parse all files + """ + + glob = GlobSourceFiles(srctree=self.config.src_tree) + + for fname in glob.parse_files(file_list, self.file_not_found_cb): + self.parse_file(fname) + + for fname in glob.parse_files(export_file, self.file_not_found_cb): + self.process_export_file(fname) + + def out_msg(self, fname, name, arg): + """ + Return output messages from a file name using the output style + filtering. + + If output type was not handled by the syler, return None. + """ + + # NOTE: we can add rules here to filter out unwanted parts, + # although OutputFormat.msg already does that. + + return self.out_style.msg(fname, name, arg) + + def msg(self, enable_lineno=False, export=False, internal=False, + symbol=None, nosymbol=None, no_doc_sections=False, + filenames=None, export_file=None): + """ + Interacts over the kernel-doc results and output messages, + returning kernel-doc markups on each interaction + """ + + self.out_style.set_config(self.config) + + if not filenames: + filenames = sorted(self.results.keys()) + + glob = GlobSourceFiles(srctree=self.config.src_tree) + + for fname in filenames: + function_table = set() + + if internal or export: + if not export_file: + export_file = [fname] + + for f in glob.parse_files(export_file, self.file_not_found_cb): + function_table |= self.export_table[f] + + if symbol: + for s in symbol: + function_table.add(s) + + self.out_style.set_filter(export, internal, symbol, nosymbol, + function_table, enable_lineno, + no_doc_sections) + + msg = "" + if fname not in self.results: + self.config.log.warning("No kernel-doc for file %s", fname) + continue + + for arg in self.results[fname]: + m = self.out_msg(fname, arg.name, arg) + + if m is None: + ln = arg.get("ln", 0) + dtype = arg.get('type', "") + + self.config.log.warning("%s:%d Can't handle %s", + fname, ln, dtype) + else: + msg += m + + if msg: + yield fname, msg diff --git a/scripts/lib/kdoc/kdoc_item.py b/scripts/lib/kdoc/kdoc_item.py new file mode 100644 index 0000000000..b3b2257645 --- /dev/null +++ b/scripts/lib/kdoc/kdoc_item.py @@ -0,0 +1,42 @@ +# SPDX-License-Identifier: GPL-2.0 +# +# A class that will, eventually, encapsulate all of the parsed data that we +# then pass into the output modules. +# + +class KdocItem: + def __init__(self, name, type, start_line, **other_stuff): + self.name = name + self.type = type + self.declaration_start_line = start_line + self.sections = {} + self.sections_start_lines = {} + self.parameterlist = [] + self.parameterdesc_start_lines = [] + self.parameterdescs = {} + self.parametertypes = {} + # + # Just save everything else into our own dict so that the output + # side can grab it directly as before. As we move things into more + # structured data, this will, hopefully, fade away. + # + self.other_stuff = other_stuff + + def get(self, key, default = None): + return self.other_stuff.get(key, default) + + def __getitem__(self, key): + return self.get(key) + + # + # Tracking of section and parameter information. + # + def set_sections(self, sections, start_lines): + self.sections = sections + self.section_start_lines = start_lines + + def set_params(self, names, descs, types, starts): + self.parameterlist = names + self.parameterdescs = descs + self.parametertypes = types + self.parameterdesc_start_lines = starts diff --git a/scripts/lib/kdoc/kdoc_output.py b/scripts/lib/kdoc/kdoc_output.py new file mode 100644 index 0000000000..ea8914537b --- /dev/null +++ b/scripts/lib/kdoc/kdoc_output.py @@ -0,0 +1,749 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +# Copyright(c) 2025: Mauro Carvalho Chehab . +# +# pylint: disable=C0301,R0902,R0911,R0912,R0913,R0914,R0915,R0917 + +""" +Implement output filters to print kernel-doc documentation. + +The implementation uses a virtual base class (OutputFormat) which +contains a dispatches to virtual methods, and some code to filter +out output messages. + +The actual implementation is done on one separate class per each type +of output. Currently, there are output classes for ReST and man/troff. +""" + +import os +import re +from datetime import datetime + +from kdoc_parser import KernelDoc, type_param +from kdoc_re import KernRe + + +function_pointer = KernRe(r"([^\(]*\(\*)\s*\)\s*\(([^\)]*)\)", cache=False) + +# match expressions used to find embedded type information +type_constant = KernRe(r"\b``([^\`]+)``\b", cache=False) +type_constant2 = KernRe(r"\%([-_*\w]+)", cache=False) +type_func = KernRe(r"(\w+)\(\)", cache=False) +type_param_ref = KernRe(r"([\!~\*]?)\@(\w*((\.\w+)|(->\w+))*(\.\.\.)?)", cache=False) + +# Special RST handling for func ptr params +type_fp_param = KernRe(r"\@(\w+)\(\)", cache=False) + +# Special RST handling for structs with func ptr params +type_fp_param2 = KernRe(r"\@(\w+->\S+)\(\)", cache=False) + +type_env = KernRe(r"(\$\w+)", cache=False) +type_enum = KernRe(r"\&(enum\s*([_\w]+))", cache=False) +type_struct = KernRe(r"\&(struct\s*([_\w]+))", cache=False) +type_typedef = KernRe(r"\&(typedef\s*([_\w]+))", cache=False) +type_union = KernRe(r"\&(union\s*([_\w]+))", cache=False) +type_member = KernRe(r"\&([_\w]+)(\.|->)([_\w]+)", cache=False) +type_fallback = KernRe(r"\&([_\w]+)", cache=False) +type_member_func = type_member + KernRe(r"\(\)", cache=False) + + +class OutputFormat: + """ + Base class for OutputFormat. If used as-is, it means that only + warnings will be displayed. + """ + + # output mode. + OUTPUT_ALL = 0 # output all symbols and doc sections + OUTPUT_INCLUDE = 1 # output only specified symbols + OUTPUT_EXPORTED = 2 # output exported symbols + OUTPUT_INTERNAL = 3 # output non-exported symbols + + # Virtual member to be overriden at the inherited classes + highlights = [] + + def __init__(self): + """Declare internal vars and set mode to OUTPUT_ALL""" + + self.out_mode = self.OUTPUT_ALL + self.enable_lineno = None + self.nosymbol = {} + self.symbol = None + self.function_table = None + self.config = None + self.no_doc_sections = False + + self.data = "" + + def set_config(self, config): + """ + Setup global config variables used by both parser and output. + """ + + self.config = config + + def set_filter(self, export, internal, symbol, nosymbol, function_table, + enable_lineno, no_doc_sections): + """ + Initialize filter variables according with the requested mode. + + Only one choice is valid between export, internal and symbol. + + The nosymbol filter can be used on all modes. + """ + + self.enable_lineno = enable_lineno + self.no_doc_sections = no_doc_sections + self.function_table = function_table + + if symbol: + self.out_mode = self.OUTPUT_INCLUDE + elif export: + self.out_mode = self.OUTPUT_EXPORTED + elif internal: + self.out_mode = self.OUTPUT_INTERNAL + else: + self.out_mode = self.OUTPUT_ALL + + if nosymbol: + self.nosymbol = set(nosymbol) + + + def highlight_block(self, block): + """ + Apply the RST highlights to a sub-block of text. + """ + + for r, sub in self.highlights: + block = r.sub(sub, block) + + return block + + def out_warnings(self, args): + """ + Output warnings for identifiers that will be displayed. + """ + + for log_msg in args.warnings: + self.config.warning(log_msg) + + def check_doc(self, name, args): + """Check if DOC should be output""" + + if self.no_doc_sections: + return False + + if name in self.nosymbol: + return False + + if self.out_mode == self.OUTPUT_ALL: + self.out_warnings(args) + return True + + if self.out_mode == self.OUTPUT_INCLUDE: + if name in self.function_table: + self.out_warnings(args) + return True + + return False + + def check_declaration(self, dtype, name, args): + """ + Checks if a declaration should be output or not based on the + filtering criteria. + """ + + if name in self.nosymbol: + return False + + if self.out_mode == self.OUTPUT_ALL: + self.out_warnings(args) + return True + + if self.out_mode in [self.OUTPUT_INCLUDE, self.OUTPUT_EXPORTED]: + if name in self.function_table: + return True + + if self.out_mode == self.OUTPUT_INTERNAL: + if dtype != "function": + self.out_warnings(args) + return True + + if name not in self.function_table: + self.out_warnings(args) + return True + + return False + + def msg(self, fname, name, args): + """ + Handles a single entry from kernel-doc parser + """ + + self.data = "" + + dtype = args.type + + if dtype == "doc": + self.out_doc(fname, name, args) + return self.data + + if not self.check_declaration(dtype, name, args): + return self.data + + if dtype == "function": + self.out_function(fname, name, args) + return self.data + + if dtype == "enum": + self.out_enum(fname, name, args) + return self.data + + if dtype == "typedef": + self.out_typedef(fname, name, args) + return self.data + + if dtype in ["struct", "union"]: + self.out_struct(fname, name, args) + return self.data + + # Warn if some type requires an output logic + self.config.log.warning("doesn't now how to output '%s' block", + dtype) + + return None + + # Virtual methods to be overridden by inherited classes + # At the base class, those do nothing. + def out_doc(self, fname, name, args): + """Outputs a DOC block""" + + def out_function(self, fname, name, args): + """Outputs a function""" + + def out_enum(self, fname, name, args): + """Outputs an enum""" + + def out_typedef(self, fname, name, args): + """Outputs a typedef""" + + def out_struct(self, fname, name, args): + """Outputs a struct""" + + +class RestFormat(OutputFormat): + """Consts and functions used by ReST output""" + + highlights = [ + (type_constant, r"``\1``"), + (type_constant2, r"``\1``"), + + # Note: need to escape () to avoid func matching later + (type_member_func, r":c:type:`\1\2\3\\(\\) <\1>`"), + (type_member, r":c:type:`\1\2\3 <\1>`"), + (type_fp_param, r"**\1\\(\\)**"), + (type_fp_param2, r"**\1\\(\\)**"), + (type_func, r"\1()"), + (type_enum, r":c:type:`\1 <\2>`"), + (type_struct, r":c:type:`\1 <\2>`"), + (type_typedef, r":c:type:`\1 <\2>`"), + (type_union, r":c:type:`\1 <\2>`"), + + # in rst this can refer to any type + (type_fallback, r":c:type:`\1`"), + (type_param_ref, r"**\1\2**") + ] + blankline = "\n" + + sphinx_literal = KernRe(r'^[^.].*::$', cache=False) + sphinx_cblock = KernRe(r'^\.\.\ +code-block::', cache=False) + + def __init__(self): + """ + Creates class variables. + + Not really mandatory, but it is a good coding style and makes + pylint happy. + """ + + super().__init__() + self.lineprefix = "" + + def print_lineno(self, ln): + """Outputs a line number""" + + if self.enable_lineno and ln is not None: + ln += 1 + self.data += f".. LINENO {ln}\n" + + def output_highlight(self, args): + """ + Outputs a C symbol that may require being converted to ReST using + the self.highlights variable + """ + + input_text = args + output = "" + in_literal = False + litprefix = "" + block = "" + + for line in input_text.strip("\n").split("\n"): + + # If we're in a literal block, see if we should drop out of it. + # Otherwise, pass the line straight through unmunged. + if in_literal: + if line.strip(): # If the line is not blank + # If this is the first non-blank line in a literal block, + # figure out the proper indent. + if not litprefix: + r = KernRe(r'^(\s*)') + if r.match(line): + litprefix = '^' + r.group(1) + else: + litprefix = "" + + output += line + "\n" + elif not KernRe(litprefix).match(line): + in_literal = False + else: + output += line + "\n" + else: + output += line + "\n" + + # Not in a literal block (or just dropped out) + if not in_literal: + block += line + "\n" + if self.sphinx_literal.match(line) or self.sphinx_cblock.match(line): + in_literal = True + litprefix = "" + output += self.highlight_block(block) + block = "" + + # Handle any remaining block + if block: + output += self.highlight_block(block) + + # Print the output with the line prefix + for line in output.strip("\n").split("\n"): + self.data += self.lineprefix + line + "\n" + + def out_section(self, args, out_docblock=False): + """ + Outputs a block section. + + This could use some work; it's used to output the DOC: sections, and + starts by putting out the name of the doc section itself, but that + tends to duplicate a header already in the template file. + """ + for section, text in args.sections.items(): + # Skip sections that are in the nosymbol_table + if section in self.nosymbol: + continue + + if out_docblock: + if not self.out_mode == self.OUTPUT_INCLUDE: + self.data += f".. _{section}:\n\n" + self.data += f'{self.lineprefix}**{section}**\n\n' + else: + self.data += f'{self.lineprefix}**{section}**\n\n' + + self.print_lineno(args.section_start_lines.get(section, 0)) + self.output_highlight(text) + self.data += "\n" + self.data += "\n" + + def out_doc(self, fname, name, args): + if not self.check_doc(name, args): + return + self.out_section(args, out_docblock=True) + + def out_function(self, fname, name, args): + + oldprefix = self.lineprefix + signature = "" + + func_macro = args.get('func_macro', False) + if func_macro: + signature = name + else: + if args.get('functiontype'): + signature = args['functiontype'] + " " + signature += name + " (" + + ln = args.declaration_start_line + count = 0 + for parameter in args.parameterlist: + if count != 0: + signature += ", " + count += 1 + dtype = args.parametertypes.get(parameter, "") + + if function_pointer.search(dtype): + signature += function_pointer.group(1) + parameter + function_pointer.group(3) + else: + signature += dtype + + if not func_macro: + signature += ")" + + self.print_lineno(ln) + if args.get('typedef') or not args.get('functiontype'): + self.data += f".. c:macro:: {name}\n\n" + + if args.get('typedef'): + self.data += " **Typedef**: " + self.lineprefix = "" + self.output_highlight(args.get('purpose', "")) + self.data += "\n\n**Syntax**\n\n" + self.data += f" ``{signature}``\n\n" + else: + self.data += f"``{signature}``\n\n" + else: + self.data += f".. c:function:: {signature}\n\n" + + if not args.get('typedef'): + self.print_lineno(ln) + self.lineprefix = " " + self.output_highlight(args.get('purpose', "")) + self.data += "\n" + + # Put descriptive text into a container (HTML
) to help set + # function prototypes apart + self.lineprefix = " " + + if args.parameterlist: + self.data += ".. container:: kernelindent\n\n" + self.data += f"{self.lineprefix}**Parameters**\n\n" + + for parameter in args.parameterlist: + parameter_name = KernRe(r'\[.*').sub('', parameter) + dtype = args.parametertypes.get(parameter, "") + + if dtype: + self.data += f"{self.lineprefix}``{dtype}``\n" + else: + self.data += f"{self.lineprefix}``{parameter}``\n" + + self.print_lineno(args.parameterdesc_start_lines.get(parameter_name, 0)) + + self.lineprefix = " " + if parameter_name in args.parameterdescs and \ + args.parameterdescs[parameter_name] != KernelDoc.undescribed: + + self.output_highlight(args.parameterdescs[parameter_name]) + self.data += "\n" + else: + self.data += f"{self.lineprefix}*undescribed*\n\n" + self.lineprefix = " " + + self.out_section(args) + self.lineprefix = oldprefix + + def out_enum(self, fname, name, args): + + oldprefix = self.lineprefix + ln = args.declaration_start_line + + self.data += f"\n\n.. c:enum:: {name}\n\n" + + self.print_lineno(ln) + self.lineprefix = " " + self.output_highlight(args.get('purpose', '')) + self.data += "\n" + + self.data += ".. container:: kernelindent\n\n" + outer = self.lineprefix + " " + self.lineprefix = outer + " " + self.data += f"{outer}**Constants**\n\n" + + for parameter in args.parameterlist: + self.data += f"{outer}``{parameter}``\n" + + if args.parameterdescs.get(parameter, '') != KernelDoc.undescribed: + self.output_highlight(args.parameterdescs[parameter]) + else: + self.data += f"{self.lineprefix}*undescribed*\n\n" + self.data += "\n" + + self.lineprefix = oldprefix + self.out_section(args) + + def out_typedef(self, fname, name, args): + + oldprefix = self.lineprefix + ln = args.declaration_start_line + + self.data += f"\n\n.. c:type:: {name}\n\n" + + self.print_lineno(ln) + self.lineprefix = " " + + self.output_highlight(args.get('purpose', '')) + + self.data += "\n" + + self.lineprefix = oldprefix + self.out_section(args) + + def out_struct(self, fname, name, args): + + purpose = args.get('purpose', "") + declaration = args.get('definition', "") + dtype = args.type + ln = args.declaration_start_line + + self.data += f"\n\n.. c:{dtype}:: {name}\n\n" + + self.print_lineno(ln) + + oldprefix = self.lineprefix + self.lineprefix += " " + + self.output_highlight(purpose) + self.data += "\n" + + self.data += ".. container:: kernelindent\n\n" + self.data += f"{self.lineprefix}**Definition**::\n\n" + + self.lineprefix = self.lineprefix + " " + + declaration = declaration.replace("\t", self.lineprefix) + + self.data += f"{self.lineprefix}{dtype} {name}" + ' {' + "\n" + self.data += f"{declaration}{self.lineprefix}" + "};\n\n" + + self.lineprefix = " " + self.data += f"{self.lineprefix}**Members**\n\n" + for parameter in args.parameterlist: + if not parameter or parameter.startswith("#"): + continue + + parameter_name = parameter.split("[", maxsplit=1)[0] + + if args.parameterdescs.get(parameter_name) == KernelDoc.undescribed: + continue + + self.print_lineno(args.parameterdesc_start_lines.get(parameter_name, 0)) + + self.data += f"{self.lineprefix}``{parameter}``\n" + + self.lineprefix = " " + self.output_highlight(args.parameterdescs[parameter_name]) + self.lineprefix = " " + + self.data += "\n" + + self.data += "\n" + + self.lineprefix = oldprefix + self.out_section(args) + + +class ManFormat(OutputFormat): + """Consts and functions used by man pages output""" + + highlights = ( + (type_constant, r"\1"), + (type_constant2, r"\1"), + (type_func, r"\\fB\1\\fP"), + (type_enum, r"\\fI\1\\fP"), + (type_struct, r"\\fI\1\\fP"), + (type_typedef, r"\\fI\1\\fP"), + (type_union, r"\\fI\1\\fP"), + (type_param, r"\\fI\1\\fP"), + (type_param_ref, r"\\fI\1\2\\fP"), + (type_member, r"\\fI\1\2\3\\fP"), + (type_fallback, r"\\fI\1\\fP") + ) + blankline = "" + + date_formats = [ + "%a %b %d %H:%M:%S %Z %Y", + "%a %b %d %H:%M:%S %Y", + "%Y-%m-%d", + "%b %d %Y", + "%B %d %Y", + "%m %d %Y", + ] + + def __init__(self, modulename): + """ + Creates class variables. + + Not really mandatory, but it is a good coding style and makes + pylint happy. + """ + + super().__init__() + self.modulename = modulename + + dt = None + tstamp = os.environ.get("KBUILD_BUILD_TIMESTAMP") + if tstamp: + for fmt in self.date_formats: + try: + dt = datetime.strptime(tstamp, fmt) + break + except ValueError: + pass + + if not dt: + dt = datetime.now() + + self.man_date = dt.strftime("%B %Y") + + def output_highlight(self, block): + """ + Outputs a C symbol that may require being highlighted with + self.highlights variable using troff syntax + """ + + contents = self.highlight_block(block) + + if isinstance(contents, list): + contents = "\n".join(contents) + + for line in contents.strip("\n").split("\n"): + line = KernRe(r"^\s*").sub("", line) + if not line: + continue + + if line[0] == ".": + self.data += "\\&" + line + "\n" + else: + self.data += line + "\n" + + def out_doc(self, fname, name, args): + if not self.check_doc(name, args): + return + + self.data += f'.TH "{self.modulename}" 9 "{self.modulename}" "{self.man_date}" "API Manual" LINUX' + "\n" + + for section, text in args.sections.items(): + self.data += f'.SH "{section}"' + "\n" + self.output_highlight(text) + + def out_function(self, fname, name, args): + """output function in man""" + + self.data += f'.TH "{name}" 9 "{name}" "{self.man_date}" "Kernel Hacker\'s Manual" LINUX' + "\n" + + self.data += ".SH NAME\n" + self.data += f"{name} \\- {args['purpose']}\n" + + self.data += ".SH SYNOPSIS\n" + if args.get('functiontype', ''): + self.data += f'.B "{args["functiontype"]}" {name}' + "\n" + else: + self.data += f'.B "{name}' + "\n" + + count = 0 + parenth = "(" + post = "," + + for parameter in args.parameterlist: + if count == len(args.parameterlist) - 1: + post = ");" + + dtype = args.parametertypes.get(parameter, "") + if function_pointer.match(dtype): + # Pointer-to-function + self.data += f'".BI "{parenth}{function_pointer.group(1)}" " ") ({function_pointer.group(2)}){post}"' + "\n" + else: + dtype = KernRe(r'([^\*])$').sub(r'\1 ', dtype) + + self.data += f'.BI "{parenth}{dtype}" "{post}"' + "\n" + count += 1 + parenth = "" + + if args.parameterlist: + self.data += ".SH ARGUMENTS\n" + + for parameter in args.parameterlist: + parameter_name = re.sub(r'\[.*', '', parameter) + + self.data += f'.IP "{parameter}" 12' + "\n" + self.output_highlight(args.parameterdescs.get(parameter_name, "")) + + for section, text in args.sections.items(): + self.data += f'.SH "{section.upper()}"' + "\n" + self.output_highlight(text) + + def out_enum(self, fname, name, args): + self.data += f'.TH "{self.modulename}" 9 "enum {name}" "{self.man_date}" "API Manual" LINUX' + "\n" + + self.data += ".SH NAME\n" + self.data += f"enum {name} \\- {args['purpose']}\n" + + self.data += ".SH SYNOPSIS\n" + self.data += f"enum {name}" + " {\n" + + count = 0 + for parameter in args.parameterlist: + self.data += f'.br\n.BI " {parameter}"' + "\n" + if count == len(args.parameterlist) - 1: + self.data += "\n};\n" + else: + self.data += ", \n.br\n" + + count += 1 + + self.data += ".SH Constants\n" + + for parameter in args.parameterlist: + parameter_name = KernRe(r'\[.*').sub('', parameter) + self.data += f'.IP "{parameter}" 12' + "\n" + self.output_highlight(args.parameterdescs.get(parameter_name, "")) + + for section, text in args.sections.items(): + self.data += f'.SH "{section}"' + "\n" + self.output_highlight(text) + + def out_typedef(self, fname, name, args): + module = self.modulename + purpose = args.get('purpose') + + self.data += f'.TH "{module}" 9 "{name}" "{self.man_date}" "API Manual" LINUX' + "\n" + + self.data += ".SH NAME\n" + self.data += f"typedef {name} \\- {purpose}\n" + + for section, text in args.sections.items(): + self.data += f'.SH "{section}"' + "\n" + self.output_highlight(text) + + def out_struct(self, fname, name, args): + module = self.modulename + purpose = args.get('purpose') + definition = args.get('definition') + + self.data += f'.TH "{module}" 9 "{args.type} {name}" "{self.man_date}" "API Manual" LINUX' + "\n" + + self.data += ".SH NAME\n" + self.data += f"{args.type} {name} \\- {purpose}\n" + + # Replace tabs with two spaces and handle newlines + declaration = definition.replace("\t", " ") + declaration = KernRe(r"\n").sub('"\n.br\n.BI "', declaration) + + self.data += ".SH SYNOPSIS\n" + self.data += f"{args.type} {name} " + "{" + "\n.br\n" + self.data += f'.BI "{declaration}\n' + "};\n.br\n\n" + + self.data += ".SH Members\n" + for parameter in args.parameterlist: + if parameter.startswith("#"): + continue + + parameter_name = re.sub(r"\[.*", "", parameter) + + if args.parameterdescs.get(parameter_name) == KernelDoc.undescribed: + continue + + self.data += f'.IP "{parameter}" 12' + "\n" + self.output_highlight(args.parameterdescs.get(parameter_name)) + + for section, text in args.sections.items(): + self.data += f'.SH "{section}"' + "\n" + self.output_highlight(text) diff --git a/scripts/lib/kdoc/kdoc_parser.py b/scripts/lib/kdoc/kdoc_parser.py new file mode 100644 index 0000000000..fe730099ec --- /dev/null +++ b/scripts/lib/kdoc/kdoc_parser.py @@ -0,0 +1,1669 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +# Copyright(c) 2025: Mauro Carvalho Chehab . +# +# pylint: disable=C0301,C0302,R0904,R0912,R0913,R0914,R0915,R0917,R1702 + +""" +kdoc_parser +=========== + +Read a C language source or header FILE and extract embedded +documentation comments +""" + +import sys +import re +from pprint import pformat + +from kdoc_re import NestedMatch, KernRe +from kdoc_item import KdocItem + +# +# Regular expressions used to parse kernel-doc markups at KernelDoc class. +# +# Let's declare them in lowercase outside any class to make easier to +# convert from the python script. +# +# As those are evaluated at the beginning, no need to cache them +# + +# Allow whitespace at end of comment start. +doc_start = KernRe(r'^/\*\*\s*$', cache=False) + +doc_end = KernRe(r'\*/', cache=False) +doc_com = KernRe(r'\s*\*\s*', cache=False) +doc_com_body = KernRe(r'\s*\* ?', cache=False) +doc_decl = doc_com + KernRe(r'(\w+)', cache=False) + +# @params and a strictly limited set of supported section names +# Specifically: +# Match @word: +# @...: +# @{section-name}: +# while trying to not match literal block starts like "example::" +# +known_section_names = 'description|context|returns?|notes?|examples?' +known_sections = KernRe(known_section_names, flags = re.I) +doc_sect = doc_com + \ + KernRe(r'\s*(\@[.\w]+|\@\.\.\.|' + known_section_names + r')\s*:([^:].*)?$', + flags=re.I, cache=False) + +doc_content = doc_com_body + KernRe(r'(.*)', cache=False) +doc_inline_start = KernRe(r'^\s*/\*\*\s*$', cache=False) +doc_inline_sect = KernRe(r'\s*\*\s*(@\s*[\w][\w\.]*\s*):(.*)', cache=False) +doc_inline_end = KernRe(r'^\s*\*/\s*$', cache=False) +doc_inline_oneline = KernRe(r'^\s*/\*\*\s*(@[\w\s]+):\s*(.*)\s*\*/\s*$', cache=False) +attribute = KernRe(r"__attribute__\s*\(\([a-z0-9,_\*\s\(\)]*\)\)", + flags=re.I | re.S, cache=False) + +export_symbol = KernRe(r'^\s*EXPORT_SYMBOL(_GPL)?\s*\(\s*(\w+)\s*\)\s*', cache=False) +export_symbol_ns = KernRe(r'^\s*EXPORT_SYMBOL_NS(_GPL)?\s*\(\s*(\w+)\s*,\s*"\S+"\)\s*', cache=False) + +type_param = KernRe(r"\@(\w*((\.\w+)|(->\w+))*(\.\.\.)?)", cache=False) + +# +# Tests for the beginning of a kerneldoc block in its various forms. +# +doc_block = doc_com + KernRe(r'DOC:\s*(.*)?', cache=False) +doc_begin_data = KernRe(r"^\s*\*?\s*(struct|union|enum|typedef)\b\s*(\w*)", cache = False) +doc_begin_func = KernRe(str(doc_com) + # initial " * ' + r"(?:\w+\s*\*\s*)?" + # type (not captured) + r'(?:define\s+)?' + # possible "define" (not captured) + r'(\w+)\s*(?:\(\w*\))?\s*' + # name and optional "(...)" + r'(?:[-:].*)?$', # description (not captured) + cache = False) + +# +# A little helper to get rid of excess white space +# +multi_space = KernRe(r'\s\s+') +def trim_whitespace(s): + return multi_space.sub(' ', s.strip()) + +class state: + """ + State machine enums + """ + + # Parser states + NORMAL = 0 # normal code + NAME = 1 # looking for function name + DECLARATION = 2 # We have seen a declaration which might not be done + BODY = 3 # the body of the comment + SPECIAL_SECTION = 4 # doc section ending with a blank line + PROTO = 5 # scanning prototype + DOCBLOCK = 6 # documentation block + INLINE_NAME = 7 # gathering doc outside main block + INLINE_TEXT = 8 # reading the body of inline docs + + name = [ + "NORMAL", + "NAME", + "DECLARATION", + "BODY", + "SPECIAL_SECTION", + "PROTO", + "DOCBLOCK", + "INLINE_NAME", + "INLINE_TEXT", + ] + + +SECTION_DEFAULT = "Description" # default section + +class KernelEntry: + + def __init__(self, config, ln): + self.config = config + + self._contents = [] + self.prototype = "" + + self.warnings = [] + + self.parameterlist = [] + self.parameterdescs = {} + self.parametertypes = {} + self.parameterdesc_start_lines = {} + + self.section_start_lines = {} + self.sections = {} + + self.anon_struct_union = False + + self.leading_space = None + + # State flags + self.brcount = 0 + self.declaration_start_line = ln + 1 + + # + # Management of section contents + # + def add_text(self, text): + self._contents.append(text) + + def contents(self): + return '\n'.join(self._contents) + '\n' + + # TODO: rename to emit_message after removal of kernel-doc.pl + def emit_msg(self, log_msg, warning=True): + """Emit a message""" + + if not warning: + self.config.log.info(log_msg) + return + + # Delegate warning output to output logic, as this way it + # will report warnings/info only for symbols that are output + + self.warnings.append(log_msg) + return + + # + # Begin a new section. + # + def begin_section(self, line_no, title = SECTION_DEFAULT, dump = False): + if dump: + self.dump_section(start_new = True) + self.section = title + self.new_start_line = line_no + + def dump_section(self, start_new=True): + """ + Dumps section contents to arrays/hashes intended for that purpose. + """ + # + # If we have accumulated no contents in the default ("description") + # section, don't bother. + # + if self.section == SECTION_DEFAULT and not self._contents: + return + name = self.section + contents = self.contents() + + if type_param.match(name): + name = type_param.group(1) + + self.parameterdescs[name] = contents + self.parameterdesc_start_lines[name] = self.new_start_line + + self.new_start_line = 0 + + else: + if name in self.sections and self.sections[name] != "": + # Only warn on user-specified duplicate section names + if name != SECTION_DEFAULT: + self.emit_msg(self.new_start_line, + f"duplicate section name '{name}'\n") + # Treat as a new paragraph - add a blank line + self.sections[name] += '\n' + contents + else: + self.sections[name] = contents + self.section_start_lines[name] = self.new_start_line + self.new_start_line = 0 + +# self.config.log.debug("Section: %s : %s", name, pformat(vars(self))) + + if start_new: + self.section = SECTION_DEFAULT + self._contents = [] + + +class KernelDoc: + """ + Read a C language source or header FILE and extract embedded + documentation comments. + """ + + # Section names + + section_context = "Context" + section_return = "Return" + + undescribed = "-- undescribed --" + + def __init__(self, config, fname): + """Initialize internal variables""" + + self.fname = fname + self.config = config + + # Initial state for the state machines + self.state = state.NORMAL + + # Store entry currently being processed + self.entry = None + + # Place all potential outputs into an array + self.entries = [] + + # + # We need Python 3.7 for its "dicts remember the insertion + # order" guarantee + # + if sys.version_info.major == 3 and sys.version_info.minor < 7: + self.emit_msg(0, + 'Python 3.7 or later is required for correct results') + + def emit_msg(self, ln, msg, warning=True): + """Emit a message""" + + log_msg = f"{self.fname}:{ln} {msg}" + + if self.entry: + self.entry.emit_msg(log_msg, warning) + return + + if warning: + self.config.log.warning(log_msg) + else: + self.config.log.info(log_msg) + + def dump_section(self, start_new=True): + """ + Dumps section contents to arrays/hashes intended for that purpose. + """ + + if self.entry: + self.entry.dump_section(start_new) + + # TODO: rename it to store_declaration after removal of kernel-doc.pl + def output_declaration(self, dtype, name, **args): + """ + Stores the entry into an entry array. + + The actual output and output filters will be handled elsewhere + """ + + item = KdocItem(name, dtype, self.entry.declaration_start_line, **args) + item.warnings = self.entry.warnings + + # Drop empty sections + # TODO: improve empty sections logic to emit warnings + sections = self.entry.sections + for section in ["Description", "Return"]: + if section in sections and not sections[section].rstrip(): + del sections[section] + item.set_sections(sections, self.entry.section_start_lines) + item.set_params(self.entry.parameterlist, self.entry.parameterdescs, + self.entry.parametertypes, + self.entry.parameterdesc_start_lines) + self.entries.append(item) + + self.config.log.debug("Output: %s:%s = %s", dtype, name, pformat(args)) + + def reset_state(self, ln): + """ + Ancillary routine to create a new entry. It initializes all + variables used by the state machine. + """ + + self.entry = KernelEntry(self.config, ln) + + # State flags + self.state = state.NORMAL + + def push_parameter(self, ln, decl_type, param, dtype, + org_arg, declaration_name): + """ + Store parameters and their descriptions at self.entry. + """ + + if self.entry.anon_struct_union and dtype == "" and param == "}": + return # Ignore the ending }; from anonymous struct/union + + self.entry.anon_struct_union = False + + param = KernRe(r'[\[\)].*').sub('', param, count=1) + + if dtype == "" and param.endswith("..."): + if KernRe(r'\w\.\.\.$').search(param): + # For named variable parameters of the form `x...`, + # remove the dots + param = param[:-3] + else: + # Handles unnamed variable parameters + param = "..." + + if param not in self.entry.parameterdescs or \ + not self.entry.parameterdescs[param]: + + self.entry.parameterdescs[param] = "variable arguments" + + elif dtype == "" and (not param or param == "void"): + param = "void" + self.entry.parameterdescs[param] = "no arguments" + + elif dtype == "" and param in ["struct", "union"]: + # Handle unnamed (anonymous) union or struct + dtype = param + param = "{unnamed_" + param + "}" + self.entry.parameterdescs[param] = "anonymous\n" + self.entry.anon_struct_union = True + + # Handle cache group enforcing variables: they do not need + # to be described in header files + elif "__cacheline_group" in param: + # Ignore __cacheline_group_begin and __cacheline_group_end + return + + # Warn if parameter has no description + # (but ignore ones starting with # as these are not parameters + # but inline preprocessor statements) + if param not in self.entry.parameterdescs and not param.startswith("#"): + self.entry.parameterdescs[param] = self.undescribed + + if "." not in param: + if decl_type == 'function': + dname = f"{decl_type} parameter" + else: + dname = f"{decl_type} member" + + self.emit_msg(ln, + f"{dname} '{param}' not described in '{declaration_name}'") + + # Strip spaces from param so that it is one continuous string on + # parameterlist. This fixes a problem where check_sections() + # cannot find a parameter like "addr[6 + 2]" because it actually + # appears as "addr[6", "+", "2]" on the parameter list. + # However, it's better to maintain the param string unchanged for + # output, so just weaken the string compare in check_sections() + # to ignore "[blah" in a parameter string. + + self.entry.parameterlist.append(param) + org_arg = KernRe(r'\s\s+').sub(' ', org_arg) + self.entry.parametertypes[param] = org_arg + + + def create_parameter_list(self, ln, decl_type, args, + splitter, declaration_name): + """ + Creates a list of parameters, storing them at self.entry. + """ + + # temporarily replace all commas inside function pointer definition + arg_expr = KernRe(r'(\([^\),]+),') + while arg_expr.search(args): + args = arg_expr.sub(r"\1#", args) + + for arg in args.split(splitter): + # Strip comments + arg = KernRe(r'\/\*.*\*\/').sub('', arg) + + # Ignore argument attributes + arg = KernRe(r'\sPOS0?\s').sub(' ', arg) + + # Strip leading/trailing spaces + arg = arg.strip() + arg = KernRe(r'\s+').sub(' ', arg, count=1) + + if arg.startswith('#'): + # Treat preprocessor directive as a typeless variable just to fill + # corresponding data structures "correctly". Catch it later in + # output_* subs. + + # Treat preprocessor directive as a typeless variable + self.push_parameter(ln, decl_type, arg, "", + "", declaration_name) + + elif KernRe(r'\(.+\)\s*\(').search(arg): + # Pointer-to-function + + arg = arg.replace('#', ',') + + r = KernRe(r'[^\(]+\(\*?\s*([\w\[\]\.]*)\s*\)') + if r.match(arg): + param = r.group(1) + else: + self.emit_msg(ln, f"Invalid param: {arg}") + param = arg + + dtype = KernRe(r'([^\(]+\(\*?)\s*' + re.escape(param)).sub(r'\1', arg) + self.push_parameter(ln, decl_type, param, dtype, + arg, declaration_name) + + elif KernRe(r'\(.+\)\s*\[').search(arg): + # Array-of-pointers + + arg = arg.replace('#', ',') + r = KernRe(r'[^\(]+\(\s*\*\s*([\w\[\]\.]*?)\s*(\s*\[\s*[\w]+\s*\]\s*)*\)') + if r.match(arg): + param = r.group(1) + else: + self.emit_msg(ln, f"Invalid param: {arg}") + param = arg + + dtype = KernRe(r'([^\(]+\(\*?)\s*' + re.escape(param)).sub(r'\1', arg) + + self.push_parameter(ln, decl_type, param, dtype, + arg, declaration_name) + + elif arg: + arg = KernRe(r'\s*:\s*').sub(":", arg) + arg = KernRe(r'\s*\[').sub('[', arg) + + args = KernRe(r'\s*,\s*').split(arg) + if args[0] and '*' in args[0]: + args[0] = re.sub(r'(\*+)\s*', r' \1', args[0]) + + first_arg = [] + r = KernRe(r'^(.*\s+)(.*?\[.*\].*)$') + if args[0] and r.match(args[0]): + args.pop(0) + first_arg.extend(r.group(1)) + first_arg.append(r.group(2)) + else: + first_arg = KernRe(r'\s+').split(args.pop(0)) + + args.insert(0, first_arg.pop()) + dtype = ' '.join(first_arg) + + for param in args: + if KernRe(r'^(\*+)\s*(.*)').match(param): + r = KernRe(r'^(\*+)\s*(.*)') + if not r.match(param): + self.emit_msg(ln, f"Invalid param: {param}") + continue + + param = r.group(1) + + self.push_parameter(ln, decl_type, r.group(2), + f"{dtype} {r.group(1)}", + arg, declaration_name) + + elif KernRe(r'(.*?):(\w+)').search(param): + r = KernRe(r'(.*?):(\w+)') + if not r.match(param): + self.emit_msg(ln, f"Invalid param: {param}") + continue + + if dtype != "": # Skip unnamed bit-fields + self.push_parameter(ln, decl_type, r.group(1), + f"{dtype}:{r.group(2)}", + arg, declaration_name) + else: + self.push_parameter(ln, decl_type, param, dtype, + arg, declaration_name) + + def check_sections(self, ln, decl_name, decl_type): + """ + Check for errors inside sections, emitting warnings if not found + parameters are described. + """ + for section in self.entry.sections: + if section not in self.entry.parameterlist and \ + not known_sections.search(section): + if decl_type == 'function': + dname = f"{decl_type} parameter" + else: + dname = f"{decl_type} member" + self.emit_msg(ln, + f"Excess {dname} '{section}' description in '{decl_name}'") + + def check_return_section(self, ln, declaration_name, return_type): + """ + If the function doesn't return void, warns about the lack of a + return description. + """ + + if not self.config.wreturn: + return + + # Ignore an empty return type (It's a macro) + # Ignore functions with a "void" return type (but not "void *") + if not return_type or KernRe(r'void\s*\w*\s*$').search(return_type): + return + + if not self.entry.sections.get("Return", None): + self.emit_msg(ln, + f"No description found for return value of '{declaration_name}'") + + def dump_struct(self, ln, proto): + """ + Store an entry for an struct or union + """ + + type_pattern = r'(struct|union)' + + qualifiers = [ + "__attribute__", + "__packed", + "__aligned", + "____cacheline_aligned_in_smp", + "____cacheline_aligned", + ] + + definition_body = r'\{(.*)\}\s*' + "(?:" + '|'.join(qualifiers) + ")?" + struct_members = KernRe(type_pattern + r'([^\{\};]+)(\{)([^\{\}]*)(\})([^\{\}\;]*)(\;)') + + # Extract struct/union definition + members = None + declaration_name = None + decl_type = None + + r = KernRe(type_pattern + r'\s+(\w+)\s*' + definition_body) + if r.search(proto): + decl_type = r.group(1) + declaration_name = r.group(2) + members = r.group(3) + else: + r = KernRe(r'typedef\s+' + type_pattern + r'\s*' + definition_body + r'\s*(\w+)\s*;') + + if r.search(proto): + decl_type = r.group(1) + declaration_name = r.group(3) + members = r.group(2) + + if not members: + self.emit_msg(ln, f"{proto} error: Cannot parse struct or union!") + return + + if self.entry.identifier != declaration_name: + self.emit_msg(ln, + f"expecting prototype for {decl_type} {self.entry.identifier}. Prototype was for {decl_type} {declaration_name} instead\n") + return + + args_pattern = r'([^,)]+)' + + sub_prefixes = [ + (KernRe(r'\/\*\s*private:.*?\/\*\s*public:.*?\*\/', re.S | re.I), ''), + (KernRe(r'\/\*\s*private:.*', re.S | re.I), ''), + + # Strip comments + (KernRe(r'\/\*.*?\*\/', re.S), ''), + + # Strip attributes + (attribute, ' '), + (KernRe(r'\s*__aligned\s*\([^;]*\)', re.S), ' '), + (KernRe(r'\s*__counted_by\s*\([^;]*\)', re.S), ' '), + (KernRe(r'\s*__counted_by_(le|be)\s*\([^;]*\)', re.S), ' '), + (KernRe(r'\s*__packed\s*', re.S), ' '), + (KernRe(r'\s*CRYPTO_MINALIGN_ATTR', re.S), ' '), + (KernRe(r'\s*____cacheline_aligned_in_smp', re.S), ' '), + (KernRe(r'\s*____cacheline_aligned', re.S), ' '), + + # Unwrap struct_group macros based on this definition: + # __struct_group(TAG, NAME, ATTRS, MEMBERS...) + # which has variants like: struct_group(NAME, MEMBERS...) + # Only MEMBERS arguments require documentation. + # + # Parsing them happens on two steps: + # + # 1. drop struct group arguments that aren't at MEMBERS, + # storing them as STRUCT_GROUP(MEMBERS) + # + # 2. remove STRUCT_GROUP() ancillary macro. + # + # The original logic used to remove STRUCT_GROUP() using an + # advanced regex: + # + # \bSTRUCT_GROUP(\(((?:(?>[^)(]+)|(?1))*)\))[^;]*; + # + # with two patterns that are incompatible with + # Python re module, as it has: + # + # - a recursive pattern: (?1) + # - an atomic grouping: (?>...) + # + # I tried a simpler version: but it didn't work either: + # \bSTRUCT_GROUP\(([^\)]+)\)[^;]*; + # + # As it doesn't properly match the end parenthesis on some cases. + # + # So, a better solution was crafted: there's now a NestedMatch + # class that ensures that delimiters after a search are properly + # matched. So, the implementation to drop STRUCT_GROUP() will be + # handled in separate. + + (KernRe(r'\bstruct_group\s*\(([^,]*,)', re.S), r'STRUCT_GROUP('), + (KernRe(r'\bstruct_group_attr\s*\(([^,]*,){2}', re.S), r'STRUCT_GROUP('), + (KernRe(r'\bstruct_group_tagged\s*\(([^,]*),([^,]*),', re.S), r'struct \1 \2; STRUCT_GROUP('), + (KernRe(r'\b__struct_group\s*\(([^,]*,){3}', re.S), r'STRUCT_GROUP('), + + # Replace macros + # + # TODO: use NestedMatch for FOO($1, $2, ...) matches + # + # it is better to also move those to the NestedMatch logic, + # to ensure that parenthesis will be properly matched. + + (KernRe(r'__ETHTOOL_DECLARE_LINK_MODE_MASK\s*\(([^\)]+)\)', re.S), r'DECLARE_BITMAP(\1, __ETHTOOL_LINK_MODE_MASK_NBITS)'), + (KernRe(r'DECLARE_PHY_INTERFACE_MASK\s*\(([^\)]+)\)', re.S), r'DECLARE_BITMAP(\1, PHY_INTERFACE_MODE_MAX)'), + (KernRe(r'DECLARE_BITMAP\s*\(' + args_pattern + r',\s*' + args_pattern + r'\)', re.S), r'unsigned long \1[BITS_TO_LONGS(\2)]'), + (KernRe(r'DECLARE_HASHTABLE\s*\(' + args_pattern + r',\s*' + args_pattern + r'\)', re.S), r'unsigned long \1[1 << ((\2) - 1)]'), + (KernRe(r'DECLARE_KFIFO\s*\(' + args_pattern + r',\s*' + args_pattern + r',\s*' + args_pattern + r'\)', re.S), r'\2 *\1'), + (KernRe(r'DECLARE_KFIFO_PTR\s*\(' + args_pattern + r',\s*' + args_pattern + r'\)', re.S), r'\2 *\1'), + (KernRe(r'(?:__)?DECLARE_FLEX_ARRAY\s*\(' + args_pattern + r',\s*' + args_pattern + r'\)', re.S), r'\1 \2[]'), + (KernRe(r'DEFINE_DMA_UNMAP_ADDR\s*\(' + args_pattern + r'\)', re.S), r'dma_addr_t \1'), + (KernRe(r'DEFINE_DMA_UNMAP_LEN\s*\(' + args_pattern + r'\)', re.S), r'__u32 \1'), + (KernRe(r'VIRTIO_DECLARE_FEATURES\s*\(' + args_pattern + r'\)', re.S), r'u64 \1; u64 \1_array[VIRTIO_FEATURES_DWORDS]'), + ] + + # Regexes here are guaranteed to have the end limiter matching + # the start delimiter. Yet, right now, only one replace group + # is allowed. + + sub_nested_prefixes = [ + (re.compile(r'\bSTRUCT_GROUP\('), r'\1'), + ] + + for search, sub in sub_prefixes: + members = search.sub(sub, members) + + nested = NestedMatch() + + for search, sub in sub_nested_prefixes: + members = nested.sub(search, sub, members) + + # Keeps the original declaration as-is + declaration = members + + # Split nested struct/union elements + # + # This loop was simpler at the original kernel-doc perl version, as + # while ($members =~ m/$struct_members/) { ... } + # reads 'members' string on each interaction. + # + # Python behavior is different: it parses 'members' only once, + # creating a list of tuples from the first interaction. + # + # On other words, this won't get nested structs. + # + # So, we need to have an extra loop on Python to override such + # re limitation. + + while True: + tuples = struct_members.findall(members) + if not tuples: + break + + for t in tuples: + newmember = "" + maintype = t[0] + s_ids = t[5] + content = t[3] + + oldmember = "".join(t) + + for s_id in s_ids.split(','): + s_id = s_id.strip() + + newmember += f"{maintype} {s_id}; " + s_id = KernRe(r'[:\[].*').sub('', s_id) + s_id = KernRe(r'^\s*\**(\S+)\s*').sub(r'\1', s_id) + + for arg in content.split(';'): + arg = arg.strip() + + if not arg: + continue + + r = KernRe(r'^([^\(]+\(\*?\s*)([\w\.]*)(\s*\).*)') + if r.match(arg): + # Pointer-to-function + dtype = r.group(1) + name = r.group(2) + extra = r.group(3) + + if not name: + continue + + if not s_id: + # Anonymous struct/union + newmember += f"{dtype}{name}{extra}; " + else: + newmember += f"{dtype}{s_id}.{name}{extra}; " + + else: + arg = arg.strip() + # Handle bitmaps + arg = KernRe(r':\s*\d+\s*').sub('', arg) + + # Handle arrays + arg = KernRe(r'\[.*\]').sub('', arg) + + # Handle multiple IDs + arg = KernRe(r'\s*,\s*').sub(',', arg) + + r = KernRe(r'(.*)\s+([\S+,]+)') + + if r.search(arg): + dtype = r.group(1) + names = r.group(2) + else: + newmember += f"{arg}; " + continue + + for name in names.split(','): + name = KernRe(r'^\s*\**(\S+)\s*').sub(r'\1', name).strip() + + if not name: + continue + + if not s_id: + # Anonymous struct/union + newmember += f"{dtype} {name}; " + else: + newmember += f"{dtype} {s_id}.{name}; " + + members = members.replace(oldmember, newmember) + + # Ignore other nested elements, like enums + members = re.sub(r'(\{[^\{\}]*\})', '', members) + + self.create_parameter_list(ln, decl_type, members, ';', + declaration_name) + self.check_sections(ln, declaration_name, decl_type) + + # Adjust declaration for better display + declaration = KernRe(r'([\{;])').sub(r'\1\n', declaration) + declaration = KernRe(r'\}\s+;').sub('};', declaration) + + # Better handle inlined enums + while True: + r = KernRe(r'(enum\s+\{[^\}]+),([^\n])') + if not r.search(declaration): + break + + declaration = r.sub(r'\1,\n\2', declaration) + + def_args = declaration.split('\n') + level = 1 + declaration = "" + for clause in def_args: + + clause = clause.strip() + clause = KernRe(r'\s+').sub(' ', clause, count=1) + + if not clause: + continue + + if '}' in clause and level > 1: + level -= 1 + + if not KernRe(r'^\s*#').match(clause): + declaration += "\t" * level + + declaration += "\t" + clause + "\n" + if "{" in clause and "}" not in clause: + level += 1 + + self.output_declaration(decl_type, declaration_name, + definition=declaration, + purpose=self.entry.declaration_purpose) + + def dump_enum(self, ln, proto): + """ + Stores an enum inside self.entries array. + """ + + # Ignore members marked private + proto = KernRe(r'\/\*\s*private:.*?\/\*\s*public:.*?\*\/', flags=re.S).sub('', proto) + proto = KernRe(r'\/\*\s*private:.*}', flags=re.S).sub('}', proto) + + # Strip comments + proto = KernRe(r'\/\*.*?\*\/', flags=re.S).sub('', proto) + + # Strip #define macros inside enums + proto = KernRe(r'#\s*((define|ifdef|if)\s+|endif)[^;]*;', flags=re.S).sub('', proto) + + # + # Parse out the name and members of the enum. Typedef form first. + # + r = KernRe(r'typedef\s+enum\s*\{(.*)\}\s*(\w*)\s*;') + if r.search(proto): + declaration_name = r.group(2) + members = r.group(1).rstrip() + # + # Failing that, look for a straight enum + # + else: + r = KernRe(r'enum\s+(\w*)\s*\{(.*)\}') + if r.match(proto): + declaration_name = r.group(1) + members = r.group(2).rstrip() + # + # OK, this isn't going to work. + # + else: + self.emit_msg(ln, f"{proto}: error: Cannot parse enum!") + return + # + # Make sure we found what we were expecting. + # + if self.entry.identifier != declaration_name: + if self.entry.identifier == "": + self.emit_msg(ln, + f"{proto}: wrong kernel-doc identifier on prototype") + else: + self.emit_msg(ln, + f"expecting prototype for enum {self.entry.identifier}. " + f"Prototype was for enum {declaration_name} instead") + return + + if not declaration_name: + declaration_name = "(anonymous)" + # + # Parse out the name of each enum member, and verify that we + # have a description for it. + # + member_set = set() + members = KernRe(r'\([^;)]*\)').sub('', members) + for arg in members.split(','): + if not arg: + continue + arg = KernRe(r'^\s*(\w+).*').sub(r'\1', arg) + self.entry.parameterlist.append(arg) + if arg not in self.entry.parameterdescs: + self.entry.parameterdescs[arg] = self.undescribed + self.emit_msg(ln, + f"Enum value '{arg}' not described in enum '{declaration_name}'") + member_set.add(arg) + # + # Ensure that every described member actually exists in the enum. + # + for k in self.entry.parameterdescs: + if k not in member_set: + self.emit_msg(ln, + f"Excess enum value '%{k}' description in '{declaration_name}'") + + self.output_declaration('enum', declaration_name, + purpose=self.entry.declaration_purpose) + + def dump_declaration(self, ln, prototype): + """ + Stores a data declaration inside self.entries array. + """ + + if self.entry.decl_type == "enum": + self.dump_enum(ln, prototype) + elif self.entry.decl_type == "typedef": + self.dump_typedef(ln, prototype) + elif self.entry.decl_type in ["union", "struct"]: + self.dump_struct(ln, prototype) + else: + # This would be a bug + self.emit_message(ln, f'Unknown declaration type: {self.entry.decl_type}') + + def dump_function(self, ln, prototype): + """ + Stores a function of function macro inside self.entries array. + """ + + func_macro = False + return_type = '' + decl_type = 'function' + + # Prefixes that would be removed + sub_prefixes = [ + (r"^static +", "", 0), + (r"^extern +", "", 0), + (r"^asmlinkage +", "", 0), + (r"^inline +", "", 0), + (r"^__inline__ +", "", 0), + (r"^__inline +", "", 0), + (r"^__always_inline +", "", 0), + (r"^noinline +", "", 0), + (r"^__FORTIFY_INLINE +", "", 0), + (r"__init +", "", 0), + (r"__init_or_module +", "", 0), + (r"__deprecated +", "", 0), + (r"__flatten +", "", 0), + (r"__meminit +", "", 0), + (r"__must_check +", "", 0), + (r"__weak +", "", 0), + (r"__sched +", "", 0), + (r"_noprof", "", 0), + (r"__printf\s*\(\s*\d*\s*,\s*\d*\s*\) +", "", 0), + (r"__(?:re)?alloc_size\s*\(\s*\d+\s*(?:,\s*\d+\s*)?\) +", "", 0), + (r"__diagnose_as\s*\(\s*\S+\s*(?:,\s*\d+\s*)*\) +", "", 0), + (r"DECL_BUCKET_PARAMS\s*\(\s*(\S+)\s*,\s*(\S+)\s*\)", r"\1, \2", 0), + (r"__attribute_const__ +", "", 0), + + # It seems that Python support for re.X is broken: + # At least for me (Python 3.13), this didn't work +# (r""" +# __attribute__\s*\(\( +# (?: +# [\w\s]+ # attribute name +# (?:\([^)]*\))? # attribute arguments +# \s*,? # optional comma at the end +# )+ +# \)\)\s+ +# """, "", re.X), + + # So, remove whitespaces and comments from it + (r"__attribute__\s*\(\((?:[\w\s]+(?:\([^)]*\))?\s*,?)+\)\)\s+", "", 0), + ] + + for search, sub, flags in sub_prefixes: + prototype = KernRe(search, flags).sub(sub, prototype) + + # Macros are a special case, as they change the prototype format + new_proto = KernRe(r"^#\s*define\s+").sub("", prototype) + if new_proto != prototype: + is_define_proto = True + prototype = new_proto + else: + is_define_proto = False + + # Yes, this truly is vile. We are looking for: + # 1. Return type (may be nothing if we're looking at a macro) + # 2. Function name + # 3. Function parameters. + # + # All the while we have to watch out for function pointer parameters + # (which IIRC is what the two sections are for), C types (these + # regexps don't even start to express all the possibilities), and + # so on. + # + # If you mess with these regexps, it's a good idea to check that + # the following functions' documentation still comes out right: + # - parport_register_device (function pointer parameters) + # - atomic_set (macro) + # - pci_match_device, __copy_to_user (long return type) + + name = r'[a-zA-Z0-9_~:]+' + prototype_end1 = r'[^\(]*' + prototype_end2 = r'[^\{]*' + prototype_end = fr'\(({prototype_end1}|{prototype_end2})\)' + + # Besides compiling, Perl qr{[\w\s]+} works as a non-capturing group. + # So, this needs to be mapped in Python with (?:...)? or (?:...)+ + + type1 = r'(?:[\w\s]+)?' + type2 = r'(?:[\w\s]+\*+)+' + + found = False + + if is_define_proto: + r = KernRe(r'^()(' + name + r')\s+') + + if r.search(prototype): + return_type = '' + declaration_name = r.group(2) + func_macro = True + + found = True + + if not found: + patterns = [ + rf'^()({name})\s*{prototype_end}', + rf'^({type1})\s+({name})\s*{prototype_end}', + rf'^({type2})\s*({name})\s*{prototype_end}', + ] + + for p in patterns: + r = KernRe(p) + + if r.match(prototype): + + return_type = r.group(1) + declaration_name = r.group(2) + args = r.group(3) + + self.create_parameter_list(ln, decl_type, args, ',', + declaration_name) + + found = True + break + if not found: + self.emit_msg(ln, + f"cannot understand function prototype: '{prototype}'") + return + + if self.entry.identifier != declaration_name: + self.emit_msg(ln, + f"expecting prototype for {self.entry.identifier}(). Prototype was for {declaration_name}() instead") + return + + self.check_sections(ln, declaration_name, "function") + + self.check_return_section(ln, declaration_name, return_type) + + if 'typedef' in return_type: + self.output_declaration(decl_type, declaration_name, + typedef=True, + functiontype=return_type, + purpose=self.entry.declaration_purpose, + func_macro=func_macro) + else: + self.output_declaration(decl_type, declaration_name, + typedef=False, + functiontype=return_type, + purpose=self.entry.declaration_purpose, + func_macro=func_macro) + + def dump_typedef(self, ln, proto): + """ + Stores a typedef inside self.entries array. + """ + + typedef_type = r'((?:\s+[\w\*]+\b){0,7}\s+(?:\w+\b|\*+))\s*' + typedef_ident = r'\*?\s*(\w\S+)\s*' + typedef_args = r'\s*\((.*)\);' + + typedef1 = KernRe(r'typedef' + typedef_type + r'\(' + typedef_ident + r'\)' + typedef_args) + typedef2 = KernRe(r'typedef' + typedef_type + typedef_ident + typedef_args) + + # Strip comments + proto = KernRe(r'/\*.*?\*/', flags=re.S).sub('', proto) + + # Parse function typedef prototypes + for r in [typedef1, typedef2]: + if not r.match(proto): + continue + + return_type = r.group(1).strip() + declaration_name = r.group(2) + args = r.group(3) + + if self.entry.identifier != declaration_name: + self.emit_msg(ln, + f"expecting prototype for typedef {self.entry.identifier}. Prototype was for typedef {declaration_name} instead\n") + return + + decl_type = 'function' + self.create_parameter_list(ln, decl_type, args, ',', declaration_name) + + self.output_declaration(decl_type, declaration_name, + typedef=True, + functiontype=return_type, + purpose=self.entry.declaration_purpose) + return + + # Handle nested parentheses or brackets + r = KernRe(r'(\(*.\)\s*|\[*.\]\s*);$') + while r.search(proto): + proto = r.sub('', proto) + + # Parse simple typedefs + r = KernRe(r'typedef.*\s+(\w+)\s*;') + if r.match(proto): + declaration_name = r.group(1) + + if self.entry.identifier != declaration_name: + self.emit_msg(ln, + f"expecting prototype for typedef {self.entry.identifier}. Prototype was for typedef {declaration_name} instead\n") + return + + self.output_declaration('typedef', declaration_name, + purpose=self.entry.declaration_purpose) + return + + self.emit_msg(ln, "error: Cannot parse typedef!") + + @staticmethod + def process_export(function_set, line): + """ + process EXPORT_SYMBOL* tags + + This method doesn't use any variable from the class, so declare it + with a staticmethod decorator. + """ + + # We support documenting some exported symbols with different + # names. A horrible hack. + suffixes = [ '_noprof' ] + + # Note: it accepts only one EXPORT_SYMBOL* per line, as having + # multiple export lines would violate Kernel coding style. + + if export_symbol.search(line): + symbol = export_symbol.group(2) + elif export_symbol_ns.search(line): + symbol = export_symbol_ns.group(2) + else: + return False + # + # Found an export, trim out any special suffixes + # + for suffix in suffixes: + # Be backward compatible with Python < 3.9 + if symbol.endswith(suffix): + symbol = symbol[:-len(suffix)] + function_set.add(symbol) + return True + + def process_normal(self, ln, line): + """ + STATE_NORMAL: looking for the /** to begin everything. + """ + + if not doc_start.match(line): + return + + # start a new entry + self.reset_state(ln) + + # next line is always the function name + self.state = state.NAME + + def process_name(self, ln, line): + """ + STATE_NAME: Looking for the "name - description" line + """ + # + # Check for a DOC: block and handle them specially. + # + if doc_block.search(line): + + if not doc_block.group(1): + self.entry.begin_section(ln, "Introduction") + else: + self.entry.begin_section(ln, doc_block.group(1)) + + self.entry.identifier = self.entry.section + self.state = state.DOCBLOCK + # + # Otherwise we're looking for a normal kerneldoc declaration line. + # + elif doc_decl.search(line): + self.entry.identifier = doc_decl.group(1) + + # Test for data declaration + if doc_begin_data.search(line): + self.entry.decl_type = doc_begin_data.group(1) + self.entry.identifier = doc_begin_data.group(2) + # + # Look for a function description + # + elif doc_begin_func.search(line): + self.entry.identifier = doc_begin_func.group(1) + self.entry.decl_type = "function" + # + # We struck out. + # + else: + self.emit_msg(ln, + f"This comment starts with '/**', but isn't a kernel-doc comment. Refer Documentation/doc-guide/kernel-doc.rst\n{line}") + self.state = state.NORMAL + return + # + # OK, set up for a new kerneldoc entry. + # + self.state = state.BODY + self.entry.identifier = self.entry.identifier.strip(" ") + # if there's no @param blocks need to set up default section here + self.entry.begin_section(ln + 1) + # + # Find the description portion, which *should* be there but + # isn't always. + # (We should be able to capture this from the previous parsing - someday) + # + r = KernRe("[-:](.*)") + if r.search(line): + self.entry.declaration_purpose = trim_whitespace(r.group(1)) + self.state = state.DECLARATION + else: + self.entry.declaration_purpose = "" + + if not self.entry.declaration_purpose and self.config.wshort_desc: + self.emit_msg(ln, + f"missing initial short description on line:\n{line}") + + if not self.entry.identifier and self.entry.decl_type != "enum": + self.emit_msg(ln, + f"wrong kernel-doc identifier on line:\n{line}") + self.state = state.NORMAL + + if self.config.verbose: + self.emit_msg(ln, + f"Scanning doc for {self.entry.decl_type} {self.entry.identifier}", + warning=False) + # + # Failed to find an identifier. Emit a warning + # + else: + self.emit_msg(ln, f"Cannot find identifier on line:\n{line}") + + # + # Helper function to determine if a new section is being started. + # + def is_new_section(self, ln, line): + if doc_sect.search(line): + self.state = state.BODY + # + # Pick out the name of our new section, tweaking it if need be. + # + newsection = doc_sect.group(1) + if newsection.lower() == 'description': + newsection = 'Description' + elif newsection.lower() == 'context': + newsection = 'Context' + self.state = state.SPECIAL_SECTION + elif newsection.lower() in ["@return", "@returns", + "return", "returns"]: + newsection = "Return" + self.state = state.SPECIAL_SECTION + elif newsection[0] == '@': + self.state = state.SPECIAL_SECTION + # + # Initialize the contents, and get the new section going. + # + newcontents = doc_sect.group(2) + if not newcontents: + newcontents = "" + self.dump_section() + self.entry.begin_section(ln, newsection) + self.entry.leading_space = None + + self.entry.add_text(newcontents.lstrip()) + return True + return False + + # + # Helper function to detect (and effect) the end of a kerneldoc comment. + # + def is_comment_end(self, ln, line): + if doc_end.search(line): + self.dump_section() + + # Look for doc_com + + doc_end: + r = KernRe(r'\s*\*\s*[a-zA-Z_0-9:\.]+\*/') + if r.match(line): + self.emit_msg(ln, f"suspicious ending line: {line}") + + self.entry.prototype = "" + self.entry.new_start_line = ln + 1 + + self.state = state.PROTO + return True + return False + + + def process_decl(self, ln, line): + """ + STATE_DECLARATION: We've seen the beginning of a declaration + """ + if self.is_new_section(ln, line) or self.is_comment_end(ln, line): + return + # + # Look for anything with the " * " line beginning. + # + if doc_content.search(line): + cont = doc_content.group(1) + # + # A blank line means that we have moved out of the declaration + # part of the comment (without any "special section" parameter + # descriptions). + # + if cont == "": + self.state = state.BODY + # + # Otherwise we have more of the declaration section to soak up. + # + else: + self.entry.declaration_purpose = \ + trim_whitespace(self.entry.declaration_purpose + ' ' + cont) + else: + # Unknown line, ignore + self.emit_msg(ln, f"bad line: {line}") + + + def process_special(self, ln, line): + """ + STATE_SPECIAL_SECTION: a section ending with a blank line + """ + # + # If we have hit a blank line (only the " * " marker), then this + # section is done. + # + if KernRe(r"\s*\*\s*$").match(line): + self.entry.begin_section(ln, dump = True) + self.state = state.BODY + return + # + # Not a blank line, look for the other ways to end the section. + # + if self.is_new_section(ln, line) or self.is_comment_end(ln, line): + return + # + # OK, we should have a continuation of the text for this section. + # + if doc_content.search(line): + cont = doc_content.group(1) + # + # If the lines of text after the first in a special section have + # leading white space, we need to trim it out or Sphinx will get + # confused. For the second line (the None case), see what we + # find there and remember it. + # + if self.entry.leading_space is None: + r = KernRe(r'^(\s+)') + if r.match(cont): + self.entry.leading_space = len(r.group(1)) + else: + self.entry.leading_space = 0 + # + # Otherwise, before trimming any leading chars, be *sure* + # that they are white space. We should maybe warn if this + # isn't the case. + # + for i in range(0, self.entry.leading_space): + if cont[i] != " ": + self.entry.leading_space = i + break + # + # Add the trimmed result to the section and we're done. + # + self.entry.add_text(cont[self.entry.leading_space:]) + else: + # Unknown line, ignore + self.emit_msg(ln, f"bad line: {line}") + + def process_body(self, ln, line): + """ + STATE_BODY: the bulk of a kerneldoc comment. + """ + if self.is_new_section(ln, line) or self.is_comment_end(ln, line): + return + + if doc_content.search(line): + cont = doc_content.group(1) + self.entry.add_text(cont) + else: + # Unknown line, ignore + self.emit_msg(ln, f"bad line: {line}") + + def process_inline_name(self, ln, line): + """STATE_INLINE_NAME: beginning of docbook comments within a prototype.""" + + if doc_inline_sect.search(line): + self.entry.begin_section(ln, doc_inline_sect.group(1)) + self.entry.add_text(doc_inline_sect.group(2).lstrip()) + self.state = state.INLINE_TEXT + elif doc_inline_end.search(line): + self.dump_section() + self.state = state.PROTO + elif doc_content.search(line): + self.emit_msg(ln, f"Incorrect use of kernel-doc format: {line}") + self.state = state.PROTO + # else ... ?? + + def process_inline_text(self, ln, line): + """STATE_INLINE_TEXT: docbook comments within a prototype.""" + + if doc_inline_end.search(line): + self.dump_section() + self.state = state.PROTO + elif doc_content.search(line): + self.entry.add_text(doc_content.group(1)) + # else ... ?? + + def syscall_munge(self, ln, proto): # pylint: disable=W0613 + """ + Handle syscall definitions + """ + + is_void = False + + # Strip newlines/CR's + proto = re.sub(r'[\r\n]+', ' ', proto) + + # Check if it's a SYSCALL_DEFINE0 + if 'SYSCALL_DEFINE0' in proto: + is_void = True + + # Replace SYSCALL_DEFINE with correct return type & function name + proto = KernRe(r'SYSCALL_DEFINE.*\(').sub('long sys_', proto) + + r = KernRe(r'long\s+(sys_.*?),') + if r.search(proto): + proto = KernRe(',').sub('(', proto, count=1) + elif is_void: + proto = KernRe(r'\)').sub('(void)', proto, count=1) + + # Now delete all of the odd-numbered commas in the proto + # so that argument types & names don't have a comma between them + count = 0 + length = len(proto) + + if is_void: + length = 0 # skip the loop if is_void + + for ix in range(length): + if proto[ix] == ',': + count += 1 + if count % 2 == 1: + proto = proto[:ix] + ' ' + proto[ix + 1:] + + return proto + + def tracepoint_munge(self, ln, proto): + """ + Handle tracepoint definitions + """ + + tracepointname = None + tracepointargs = None + + # Match tracepoint name based on different patterns + r = KernRe(r'TRACE_EVENT\((.*?),') + if r.search(proto): + tracepointname = r.group(1) + + r = KernRe(r'DEFINE_SINGLE_EVENT\((.*?),') + if r.search(proto): + tracepointname = r.group(1) + + r = KernRe(r'DEFINE_EVENT\((.*?),(.*?),') + if r.search(proto): + tracepointname = r.group(2) + + if tracepointname: + tracepointname = tracepointname.lstrip() + + r = KernRe(r'TP_PROTO\((.*?)\)') + if r.search(proto): + tracepointargs = r.group(1) + + if not tracepointname or not tracepointargs: + self.emit_msg(ln, + f"Unrecognized tracepoint format:\n{proto}\n") + else: + proto = f"static inline void trace_{tracepointname}({tracepointargs})" + self.entry.identifier = f"trace_{self.entry.identifier}" + + return proto + + def process_proto_function(self, ln, line): + """Ancillary routine to process a function prototype""" + + # strip C99-style comments to end of line + line = KernRe(r"\/\/.*$", re.S).sub('', line) + # + # Soak up the line's worth of prototype text, stopping at { or ; if present. + # + if KernRe(r'\s*#\s*define').match(line): + self.entry.prototype = line + elif not line.startswith('#'): # skip other preprocessor stuff + r = KernRe(r'([^\{]*)') + if r.match(line): + self.entry.prototype += r.group(1) + " " + # + # If we now have the whole prototype, clean it up and declare victory. + # + if '{' in line or ';' in line or KernRe(r'\s*#\s*define').match(line): + # strip comments and surrounding spaces + self.entry.prototype = KernRe(r'/\*.*\*/').sub('', self.entry.prototype).strip() + # + # Handle self.entry.prototypes for function pointers like: + # int (*pcs_config)(struct foo) + # by turning it into + # int pcs_config(struct foo) + # + r = KernRe(r'^(\S+\s+)\(\s*\*(\S+)\)') + self.entry.prototype = r.sub(r'\1\2', self.entry.prototype) + # + # Handle special declaration syntaxes + # + if 'SYSCALL_DEFINE' in self.entry.prototype: + self.entry.prototype = self.syscall_munge(ln, + self.entry.prototype) + else: + r = KernRe(r'TRACE_EVENT|DEFINE_EVENT|DEFINE_SINGLE_EVENT') + if r.search(self.entry.prototype): + self.entry.prototype = self.tracepoint_munge(ln, + self.entry.prototype) + # + # ... and we're done + # + self.dump_function(ln, self.entry.prototype) + self.reset_state(ln) + + def process_proto_type(self, ln, line): + """Ancillary routine to process a type""" + + # Strip C99-style comments and surrounding whitespace + line = KernRe(r"//.*$", re.S).sub('', line).strip() + if not line: + return # nothing to see here + + # To distinguish preprocessor directive from regular declaration later. + if line.startswith('#'): + line += ";" + # + # Split the declaration on any of { } or ;, and accumulate pieces + # until we hit a semicolon while not inside {brackets} + # + r = KernRe(r'(.*?)([{};])') + for chunk in r.split(line): + if chunk: # Ignore empty matches + self.entry.prototype += chunk + # + # This cries out for a match statement ... someday after we can + # drop Python 3.9 ... + # + if chunk == '{': + self.entry.brcount += 1 + elif chunk == '}': + self.entry.brcount -= 1 + elif chunk == ';' and self.entry.brcount <= 0: + self.dump_declaration(ln, self.entry.prototype) + self.reset_state(ln) + return + # + # We hit the end of the line while still in the declaration; put + # in a space to represent the newline. + # + self.entry.prototype += ' ' + + def process_proto(self, ln, line): + """STATE_PROTO: reading a function/whatever prototype.""" + + if doc_inline_oneline.search(line): + self.entry.begin_section(ln, doc_inline_oneline.group(1)) + self.entry.add_text(doc_inline_oneline.group(2)) + self.dump_section() + + elif doc_inline_start.search(line): + self.state = state.INLINE_NAME + + elif self.entry.decl_type == 'function': + self.process_proto_function(ln, line) + + else: + self.process_proto_type(ln, line) + + def process_docblock(self, ln, line): + """STATE_DOCBLOCK: within a DOC: block.""" + + if doc_end.search(line): + self.dump_section() + self.output_declaration("doc", self.entry.identifier) + self.reset_state(ln) + + elif doc_content.search(line): + self.entry.add_text(doc_content.group(1)) + + def parse_export(self): + """ + Parses EXPORT_SYMBOL* macros from a single Kernel source file. + """ + + export_table = set() + + try: + with open(self.fname, "r", encoding="utf8", + errors="backslashreplace") as fp: + + for line in fp: + self.process_export(export_table, line) + + except IOError: + return None + + return export_table + + # + # The state/action table telling us which function to invoke in + # each state. + # + state_actions = { + state.NORMAL: process_normal, + state.NAME: process_name, + state.BODY: process_body, + state.DECLARATION: process_decl, + state.SPECIAL_SECTION: process_special, + state.INLINE_NAME: process_inline_name, + state.INLINE_TEXT: process_inline_text, + state.PROTO: process_proto, + state.DOCBLOCK: process_docblock, + } + + def parse_kdoc(self): + """ + Open and process each line of a C source file. + The parsing is controlled via a state machine, and the line is passed + to a different process function depending on the state. The process + function may update the state as needed. + + Besides parsing kernel-doc tags, it also parses export symbols. + """ + + prev = "" + prev_ln = None + export_table = set() + + try: + with open(self.fname, "r", encoding="utf8", + errors="backslashreplace") as fp: + for ln, line in enumerate(fp): + + line = line.expandtabs().strip("\n") + + # Group continuation lines on prototypes + if self.state == state.PROTO: + if line.endswith("\\"): + prev += line.rstrip("\\") + if not prev_ln: + prev_ln = ln + continue + + if prev: + ln = prev_ln + line = prev + line + prev = "" + prev_ln = None + + self.config.log.debug("%d %s: %s", + ln, state.name[self.state], + line) + + # This is an optimization over the original script. + # There, when export_file was used for the same file, + # it was read twice. Here, we use the already-existing + # loop to parse exported symbols as well. + # + if (self.state != state.NORMAL) or \ + not self.process_export(export_table, line): + # Hand this line to the appropriate state handler + self.state_actions[self.state](self, ln, line) + + except OSError: + self.config.log.error(f"Error: Cannot open file {self.fname}") + + return export_table, self.entries diff --git a/scripts/lib/kdoc/kdoc_re.py b/scripts/lib/kdoc/kdoc_re.py new file mode 100644 index 0000000000..612223e1e7 --- /dev/null +++ b/scripts/lib/kdoc/kdoc_re.py @@ -0,0 +1,270 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: GPL-2.0 +# Copyright(c) 2025: Mauro Carvalho Chehab . + +""" +Regular expression ancillary classes. + +Those help caching regular expressions and do matching for kernel-doc. +""" + +import re + +# Local cache for regular expressions +re_cache = {} + + +class KernRe: + """ + Helper class to simplify regex declaration and usage, + + It calls re.compile for a given pattern. It also allows adding + regular expressions and define sub at class init time. + + Regular expressions can be cached via an argument, helping to speedup + searches. + """ + + def _add_regex(self, string, flags): + """ + Adds a new regex or re-use it from the cache. + """ + self.regex = re_cache.get(string, None) + if not self.regex: + self.regex = re.compile(string, flags=flags) + if self.cache: + re_cache[string] = self.regex + + def __init__(self, string, cache=True, flags=0): + """ + Compile a regular expression and initialize internal vars. + """ + + self.cache = cache + self.last_match = None + + self._add_regex(string, flags) + + def __str__(self): + """ + Return the regular expression pattern. + """ + return self.regex.pattern + + def __add__(self, other): + """ + Allows adding two regular expressions into one. + """ + + return KernRe(str(self) + str(other), cache=self.cache or other.cache, + flags=self.regex.flags | other.regex.flags) + + def match(self, string): + """ + Handles a re.match storing its results + """ + + self.last_match = self.regex.match(string) + return self.last_match + + def search(self, string): + """ + Handles a re.search storing its results + """ + + self.last_match = self.regex.search(string) + return self.last_match + + def findall(self, string): + """ + Alias to re.findall + """ + + return self.regex.findall(string) + + def split(self, string): + """ + Alias to re.split + """ + + return self.regex.split(string) + + def sub(self, sub, string, count=0): + """ + Alias to re.sub + """ + + return self.regex.sub(sub, string, count=count) + + def group(self, num): + """ + Returns the group results of the last match + """ + + return self.last_match.group(num) + + +class NestedMatch: + """ + Finding nested delimiters is hard with regular expressions. It is + even harder on Python with its normal re module, as there are several + advanced regular expressions that are missing. + + This is the case of this pattern: + + '\\bSTRUCT_GROUP(\\(((?:(?>[^)(]+)|(?1))*)\\))[^;]*;' + + which is used to properly match open/close parenthesis of the + string search STRUCT_GROUP(), + + Add a class that counts pairs of delimiters, using it to match and + replace nested expressions. + + The original approach was suggested by: + https://stackoverflow.com/questions/5454322/python-how-to-match-nested-parentheses-with-regex + + Although I re-implemented it to make it more generic and match 3 types + of delimiters. The logic checks if delimiters are paired. If not, it + will ignore the search string. + """ + + # TODO: make NestedMatch handle multiple match groups + # + # Right now, regular expressions to match it are defined only up to + # the start delimiter, e.g.: + # + # \bSTRUCT_GROUP\( + # + # is similar to: STRUCT_GROUP\((.*)\) + # except that the content inside the match group is delimiter's aligned. + # + # The content inside parenthesis are converted into a single replace + # group (e.g. r`\1'). + # + # It would be nice to change such definition to support multiple + # match groups, allowing a regex equivalent to. + # + # FOO\((.*), (.*), (.*)\) + # + # it is probably easier to define it not as a regular expression, but + # with some lexical definition like: + # + # FOO(arg1, arg2, arg3) + + DELIMITER_PAIRS = { + '{': '}', + '(': ')', + '[': ']', + } + + RE_DELIM = re.compile(r'[\{\}\[\]\(\)]') + + def _search(self, regex, line): + """ + Finds paired blocks for a regex that ends with a delimiter. + + The suggestion of using finditer to match pairs came from: + https://stackoverflow.com/questions/5454322/python-how-to-match-nested-parentheses-with-regex + but I ended using a different implementation to align all three types + of delimiters and seek for an initial regular expression. + + The algorithm seeks for open/close paired delimiters and place them + into a stack, yielding a start/stop position of each match when the + stack is zeroed. + + The algorithm shoud work fine for properly paired lines, but will + silently ignore end delimiters that preceeds an start delimiter. + This should be OK for kernel-doc parser, as unaligned delimiters + would cause compilation errors. So, we don't need to rise exceptions + to cover such issues. + """ + + stack = [] + + for match_re in regex.finditer(line): + start = match_re.start() + offset = match_re.end() + + d = line[offset - 1] + if d not in self.DELIMITER_PAIRS: + continue + + end = self.DELIMITER_PAIRS[d] + stack.append(end) + + for match in self.RE_DELIM.finditer(line[offset:]): + pos = match.start() + offset + + d = line[pos] + + if d in self.DELIMITER_PAIRS: + end = self.DELIMITER_PAIRS[d] + + stack.append(end) + continue + + # Does the end delimiter match what it is expected? + if stack and d == stack[-1]: + stack.pop() + + if not stack: + yield start, offset, pos + 1 + break + + def search(self, regex, line): + """ + This is similar to re.search: + + It matches a regex that it is followed by a delimiter, + returning occurrences only if all delimiters are paired. + """ + + for t in self._search(regex, line): + + yield line[t[0]:t[2]] + + def sub(self, regex, sub, line, count=0): + """ + This is similar to re.sub: + + It matches a regex that it is followed by a delimiter, + replacing occurrences only if all delimiters are paired. + + if r'\1' is used, it works just like re: it places there the + matched paired data with the delimiter stripped. + + If count is different than zero, it will replace at most count + items. + """ + out = "" + + cur_pos = 0 + n = 0 + + for start, end, pos in self._search(regex, line): + out += line[cur_pos:start] + + # Value, ignoring start/end delimiters + value = line[end:pos - 1] + + # replaces \1 at the sub string, if \1 is used there + new_sub = sub + new_sub = new_sub.replace(r'\1', value) + + out += new_sub + + # Drop end ';' if any + if line[pos] == ';': + pos += 1 + + cur_pos = pos + n += 1 + + if count and count >= n: + break + + # Append the remaining string + l = len(line) + out += line[cur_pos:l] + + return out From 7f58572cff7e79c7733ba504cd200af2287706f5 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:19 +0100 Subject: [PATCH 19/32] scripts/kernel-doc: strip QEMU_ from function definitions This commit is the Python version of our older commit b30df2751e5 ("scripts/kernel-doc: strip QEMU_ from function definitions"). Some versions of Sphinx get confused if function attributes are left on the C code from kernel-doc; strip out any QEMU_* prefixes from function prototypes. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-5-peter.maydell@linaro.org --- scripts/lib/kdoc/kdoc_parser.py | 1 + 1 file changed, 1 insertion(+) diff --git a/scripts/lib/kdoc/kdoc_parser.py b/scripts/lib/kdoc/kdoc_parser.py index fe730099ec..32b4356292 100644 --- a/scripts/lib/kdoc/kdoc_parser.py +++ b/scripts/lib/kdoc/kdoc_parser.py @@ -907,6 +907,7 @@ class KernelDoc: (r"^__always_inline +", "", 0), (r"^noinline +", "", 0), (r"^__FORTIFY_INLINE +", "", 0), + (r"QEMU_[A-Z_]+ +", "", 0), (r"__init +", "", 0), (r"__init_or_module +", "", 0), (r"__deprecated +", "", 0), From 9cbe72b868b7cb04f7553220738f1286b6f0dc32 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:20 +0100 Subject: [PATCH 20/32] scripts/kernel-doc: tweak for QEMU coding standards This commit makes the equivalent changes to the Python script that we had for the old Perl script in commit 4cf41794411f ("docs: tweak kernel-doc for QEMU coding standards"). To repeat the rationale from that commit: Surprisingly, QEMU does have a pretty consistent doc comment style and it is not very different from the Linux kernel's. Of the documentation "sigils", only "#" separates the QEMU doc comment style from Linux's, and it has 200+ instances vs. 6 for the kernel's '&struct foo' (all in accel/tcg/translate-all.c), so it's clear that the two standards are different in this respect. In addition, our structs are typedefed and recognized by CamelCase names. Note that in 4cf41794411f we used '(?!)' as our type_fallback regex; this is strictly not quite a replacement for the upstream '\&([_\w]+)', because the latter includes a group that can later be matched with \1, and the former does not. The old perl script did not care about this, but the python version does, so we must include the extra set of brackets to ensure we have a group. This commit does not include all the same changes that 4cf41794411f did. Of the missing pieces, some had already gone in an earlier kernel-doc update; the parts we still had but do not include here are: @@ -2057,7 +2060,7 @@ } elsif (/$doc_decl/o) { $identifier = $1; - if (/\s*([\w\s]+?)(\(\))?\s*-/) { + if (/\s*([\w\s]+?)(\s*-|:)/) { $identifier = $1; } @@ -2067,7 +2070,7 @@ $contents = ""; $section = $section_default; $new_start_line = $. + 1; - if (/-(.*)/) { + if (/[-:](.*)/) { # strip leading/trailing/multiple spaces $descr= $1; $descr =~ s/^\s*//; The second of these is already in the upstream version: the line r = KernRe("[-:](.*)") in process_name() matches the regex we have. The first change has been refactored into the doc_begin_data and doc_begin_func changes. Since the output HTML for QEMU's documentation has no relevant changes with the new kerneldoc, we assume that this too has been handled upstream. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-6-peter.maydell@linaro.org --- scripts/lib/kdoc/kdoc_output.py | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/scripts/lib/kdoc/kdoc_output.py b/scripts/lib/kdoc/kdoc_output.py index ea8914537b..39fa872dfc 100644 --- a/scripts/lib/kdoc/kdoc_output.py +++ b/scripts/lib/kdoc/kdoc_output.py @@ -38,12 +38,12 @@ type_fp_param = KernRe(r"\@(\w+)\(\)", cache=False) type_fp_param2 = KernRe(r"\@(\w+->\S+)\(\)", cache=False) type_env = KernRe(r"(\$\w+)", cache=False) -type_enum = KernRe(r"\&(enum\s*([_\w]+))", cache=False) -type_struct = KernRe(r"\&(struct\s*([_\w]+))", cache=False) -type_typedef = KernRe(r"\&(typedef\s*([_\w]+))", cache=False) -type_union = KernRe(r"\&(union\s*([_\w]+))", cache=False) -type_member = KernRe(r"\&([_\w]+)(\.|->)([_\w]+)", cache=False) -type_fallback = KernRe(r"\&([_\w]+)", cache=False) +type_enum = KernRe(r"#(enum\s*([_\w]+))", cache=False) +type_struct = KernRe(r"#(struct\s*([_\w]+))", cache=False) +type_typedef = KernRe(r"#(([A-Z][_\w]*))", cache=False) +type_union = KernRe(r"#(union\s*([_\w]+))", cache=False) +type_member = KernRe(r"#([_\w]+)(\.|->)([_\w]+)", cache=False) +type_fallback = KernRe(r"((?!))", cache=False) # this never matches type_member_func = type_member + KernRe(r"\(\)", cache=False) From 33be8171e2fe02173d24ffd61bd97bf9c2b37834 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:21 +0100 Subject: [PATCH 21/32] scripts/kerneldoc: Switch to the Python kernel-doc script Change the Sphinx config to run the new Python kernel-doc script instead of the Perl one. The only difference between the two is that the new script does not handle the -sphinx-version option, instead assuming that Sphinx is always at least version 3: so we must delete the code that passes that option to avoid the Python script complaining about an unknown option. QEMU's minimum Sphinx version is already 3.4.3, so this doesn't change the set of versions we can handle. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-7-peter.maydell@linaro.org --- docs/conf.py | 4 +++- docs/sphinx/kerneldoc.py | 5 ----- 2 files changed, 3 insertions(+), 6 deletions(-) diff --git a/docs/conf.py b/docs/conf.py index f892a6e1da..e09769e5f8 100644 --- a/docs/conf.py +++ b/docs/conf.py @@ -341,7 +341,9 @@ man_make_section_directory = False # We use paths starting from qemu_docdir here so that you can run # sphinx-build from anywhere and the kerneldoc extension can still # find everything. -kerneldoc_bin = ['perl', os.path.join(qemu_docdir, '../scripts/kernel-doc')] +# Since kernel-doc is now a Python script, we should run it with whatever +# Python this sphinx is using (rather than letting it find one via env) +kerneldoc_bin = [sys.executable, os.path.join(qemu_docdir, '../scripts/kernel-doc.py')] kerneldoc_srctree = os.path.join(qemu_docdir, '..') hxtool_srctree = os.path.join(qemu_docdir, '..') qapidoc_srctree = os.path.join(qemu_docdir, '..') diff --git a/docs/sphinx/kerneldoc.py b/docs/sphinx/kerneldoc.py index 30bb343198..9721072e47 100644 --- a/docs/sphinx/kerneldoc.py +++ b/docs/sphinx/kerneldoc.py @@ -63,11 +63,6 @@ class KernelDocDirective(Directive): env = self.state.document.settings.env cmd = env.config.kerneldoc_bin + ['-rst', '-enable-lineno'] - # Pass the version string to kernel-doc, as it needs to use a different - # dialect, depending what the C domain supports for each specific - # Sphinx versions - cmd += ['-sphinx-version', sphinx.__version__] - # Pass through the warnings-as-errors flag if env.config.kerneldoc_werror: cmd += ['-Werror'] From 619d5f0211ac69d71291505a8528671ab83764e3 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:22 +0100 Subject: [PATCH 22/32] scripts/kernel-doc: Delete the old Perl kernel-doc script We can now delete the old Perl kernel-doc script. For posterity, this is a complete diff of the local changes that we were carrying between the kernel's Perl script as of kernel commit 72b97d0b911872ba (the last time we synced it) and our local copy: --- /tmp/kdoc 2025-08-14 10:42:47.620331939 +0100 +++ scripts/kernel-doc 2025-02-17 10:44:34.528421457 +0000 @@ -1,5 +1,5 @@ #!/usr/bin/env perl -# SPDX-License-Identifier: GPL-2.0 +# SPDX-License-Identifier: GPL-2.0-only use warnings; use strict; @@ -224,12 +224,12 @@ my $type_fp_param = '\@(\w+)\(\)'; # Special RST handling for func ptr params my $type_fp_param2 = '\@(\w+->\S+)\(\)'; # Special RST handling for structs with func ptr params my $type_env = '(\$\w+)'; -my $type_enum = '\&(enum\s*([_\w]+))'; -my $type_struct = '\&(struct\s*([_\w]+))'; -my $type_typedef = '\&(typedef\s*([_\w]+))'; -my $type_union = '\&(union\s*([_\w]+))'; -my $type_member = '\&([_\w]+)(\.|->)([_\w]+)'; -my $type_fallback = '\&([_\w]+)'; +my $type_enum = '#(enum\s*([_\w]+))'; +my $type_struct = '#(struct\s*([_\w]+))'; +my $type_typedef = '#(([A-Z][_\w]*))'; +my $type_union = '#(union\s*([_\w]+))'; +my $type_member = '#([_\w]+)(\.|->)([_\w]+)'; +my $type_fallback = '(?!)'; # this never matches my $type_member_func = $type_member . '\(\)'; # Output conversion substitutions. @@ -1745,6 +1745,9 @@ )+ \)\)\s+//x; + # Strip QEMU specific compiler annotations + $prototype =~ s/QEMU_[A-Z_]+ +//; + # Yes, this truly is vile. We are looking for: # 1. Return type (may be nothing if we're looking at a macro) # 2. Function name @@ -2057,7 +2060,7 @@ } elsif (/$doc_decl/o) { $identifier = $1; - if (/\s*([\w\s]+?)(\(\))?\s*-/) { + if (/\s*([\w\s]+?)(\s*-|:)/) { $identifier = $1; } @@ -2067,7 +2070,7 @@ $contents = ""; $section = $section_default; $new_start_line = $. + 1; - if (/-(.*)/) { + if (/[-:](.*)/) { # strip leading/trailing/multiple spaces $descr= $1; $descr =~ s/^\s*//; These changes correspond to: 06e2329636f license: Update deprecated SPDX tag GPL-2.0 to GPL-2.0-only (a bulk change which we won't bother to re-apply to this third-party script) b30df2751e5 scripts/kernel-doc: strip QEMU_ from function definitions 4cf41794411 docs: tweak kernel-doc for QEMU coding standards We have already applied the equivalent of these changes to the Python code in libs/kdoc/ in the preceding commits. Signed-off-by: Peter Maydell Reviewed-by: Mauro Carvalho Chehab Reviewed-by: Paolo Bonzini --- .editorconfig | 2 +- scripts/kernel-doc | 2441 -------------------------------------------- 2 files changed, 1 insertion(+), 2442 deletions(-) delete mode 100755 scripts/kernel-doc diff --git a/.editorconfig b/.editorconfig index a04cb9054c..258d41ab48 100644 --- a/.editorconfig +++ b/.editorconfig @@ -55,7 +55,7 @@ indent_size = 4 emacs_mode = perl # but user kernel "style" for imported scripts -[scripts/{kernel-doc,get_maintainer.pl,checkpatch.pl}] +[scripts/{get_maintainer.pl,checkpatch.pl}] indent_style = tab indent_size = 8 emacs_mode = perl diff --git a/scripts/kernel-doc b/scripts/kernel-doc deleted file mode 100755 index 117ec8fcd1..0000000000 --- a/scripts/kernel-doc +++ /dev/null @@ -1,2441 +0,0 @@ -#!/usr/bin/env perl -# SPDX-License-Identifier: GPL-2.0-only - -use warnings; -use strict; - -## Copyright (c) 1998 Michael Zucchi, All Rights Reserved ## -## Copyright (C) 2000, 1 Tim Waugh ## -## Copyright (C) 2001 Simon Huggins ## -## Copyright (C) 2005-2012 Randy Dunlap ## -## Copyright (C) 2012 Dan Luedtke ## -## ## -## #define enhancements by Armin Kuster ## -## Copyright (c) 2000 MontaVista Software, Inc. ## -## ## -## This software falls under the GNU General Public License. ## -## Please read the COPYING file for more information ## - -# 18/01/2001 - Cleanups -# Functions prototyped as foo(void) same as foo() -# Stop eval'ing where we don't need to. -# -- huggie@earth.li - -# 27/06/2001 - Allowed whitespace after initial "/**" and -# allowed comments before function declarations. -# -- Christian Kreibich - -# Still to do: -# - add perldoc documentation -# - Look more closely at some of the scarier bits :) - -# 26/05/2001 - Support for separate source and object trees. -# Return error code. -# Keith Owens - -# 23/09/2001 - Added support for typedefs, structs, enums and unions -# Support for Context section; can be terminated using empty line -# Small fixes (like spaces vs. \s in regex) -# -- Tim Jansen - -# 25/07/2012 - Added support for HTML5 -# -- Dan Luedtke - -sub usage { - my $message = <<"EOF"; -Usage: $0 [OPTION ...] FILE ... - -Read C language source or header FILEs, extract embedded documentation comments, -and print formatted documentation to standard output. - -The documentation comments are identified by "/**" opening comment mark. See -Documentation/doc-guide/kernel-doc.rst for the documentation comment syntax. - -Output format selection (mutually exclusive): - -man Output troff manual page format. This is the default. - -rst Output reStructuredText format. - -none Do not output documentation, only warnings. - -Output format selection modifier (affects only ReST output): - - -sphinx-version Use the ReST C domain dialect compatible with an - specific Sphinx Version. - If not specified, kernel-doc will auto-detect using - the sphinx-build version found on PATH. - -Output selection (mutually exclusive): - -export Only output documentation for symbols that have been - exported using EXPORT_SYMBOL() or EXPORT_SYMBOL_GPL() - in any input FILE or -export-file FILE. - -internal Only output documentation for symbols that have NOT been - exported using EXPORT_SYMBOL() or EXPORT_SYMBOL_GPL() - in any input FILE or -export-file FILE. - -function NAME Only output documentation for the given function(s) - or DOC: section title(s). All other functions and DOC: - sections are ignored. May be specified multiple times. - -nosymbol NAME Exclude the specified symbols from the output - documentation. May be specified multiple times. - -Output selection modifiers: - -no-doc-sections Do not output DOC: sections. - -enable-lineno Enable output of #define LINENO lines. Only works with - reStructuredText format. - -export-file FILE Specify an additional FILE in which to look for - EXPORT_SYMBOL() and EXPORT_SYMBOL_GPL(). To be used with - -export or -internal. May be specified multiple times. - -Other parameters: - -v Verbose output, more warnings and other information. - -h Print this help. - -Werror Treat warnings as errors. - -EOF - print $message; - exit 1; -} - -# -# format of comments. -# In the following table, (...)? signifies optional structure. -# (...)* signifies 0 or more structure elements -# /** -# * function_name(:)? (- short description)? -# (* @parameterx: (description of parameter x)?)* -# (* a blank line)? -# * (Description:)? (Description of function)? -# * (section header: (section description)? )* -# (*)?*/ -# -# So .. the trivial example would be: -# -# /** -# * my_function -# */ -# -# If the Description: header tag is omitted, then there must be a blank line -# after the last parameter specification. -# e.g. -# /** -# * my_function - does my stuff -# * @my_arg: its mine damnit -# * -# * Does my stuff explained. -# */ -# -# or, could also use: -# /** -# * my_function - does my stuff -# * @my_arg: its mine damnit -# * Description: Does my stuff explained. -# */ -# etc. -# -# Besides functions you can also write documentation for structs, unions, -# enums and typedefs. Instead of the function name you must write the name -# of the declaration; the struct/union/enum/typedef must always precede -# the name. Nesting of declarations is not supported. -# Use the argument mechanism to document members or constants. -# e.g. -# /** -# * struct my_struct - short description -# * @a: first member -# * @b: second member -# * -# * Longer description -# */ -# struct my_struct { -# int a; -# int b; -# /* private: */ -# int c; -# }; -# -# All descriptions can be multiline, except the short function description. -# -# For really longs structs, you can also describe arguments inside the -# body of the struct. -# eg. -# /** -# * struct my_struct - short description -# * @a: first member -# * @b: second member -# * -# * Longer description -# */ -# struct my_struct { -# int a; -# int b; -# /** -# * @c: This is longer description of C -# * -# * You can use paragraphs to describe arguments -# * using this method. -# */ -# int c; -# }; -# -# This should be use only for struct/enum members. -# -# You can also add additional sections. When documenting kernel functions you -# should document the "Context:" of the function, e.g. whether the functions -# can be called form interrupts. Unlike other sections you can end it with an -# empty line. -# A non-void function should have a "Return:" section describing the return -# value(s). -# Example-sections should contain the string EXAMPLE so that they are marked -# appropriately in DocBook. -# -# Example: -# /** -# * user_function - function that can only be called in user context -# * @a: some argument -# * Context: !in_interrupt() -# * -# * Some description -# * Example: -# * user_function(22); -# */ -# ... -# -# -# All descriptive text is further processed, scanning for the following special -# patterns, which are highlighted appropriately. -# -# 'funcname()' - function -# '$ENVVAR' - environmental variable -# '&struct_name' - name of a structure (up to two words including 'struct') -# '&struct_name.member' - name of a structure member -# '@parameter' - name of a parameter -# '%CONST' - name of a constant. -# '``LITERAL``' - literal string without any spaces on it. - -## init lots of data - -my $errors = 0; -my $warnings = 0; -my $anon_struct_union = 0; - -# match expressions used to find embedded type information -my $type_constant = '\b``([^\`]+)``\b'; -my $type_constant2 = '\%([-_\w]+)'; -my $type_func = '(\w+)\(\)'; -my $type_param = '\@(\w*((\.\w+)|(->\w+))*(\.\.\.)?)'; -my $type_param_ref = '([\!]?)\@(\w*((\.\w+)|(->\w+))*(\.\.\.)?)'; -my $type_fp_param = '\@(\w+)\(\)'; # Special RST handling for func ptr params -my $type_fp_param2 = '\@(\w+->\S+)\(\)'; # Special RST handling for structs with func ptr params -my $type_env = '(\$\w+)'; -my $type_enum = '#(enum\s*([_\w]+))'; -my $type_struct = '#(struct\s*([_\w]+))'; -my $type_typedef = '#(([A-Z][_\w]*))'; -my $type_union = '#(union\s*([_\w]+))'; -my $type_member = '#([_\w]+)(\.|->)([_\w]+)'; -my $type_fallback = '(?!)'; # this never matches -my $type_member_func = $type_member . '\(\)'; - -# Output conversion substitutions. -# One for each output format - -# these are pretty rough -my @highlights_man = ( - [$type_constant, "\$1"], - [$type_constant2, "\$1"], - [$type_func, "\\\\fB\$1\\\\fP"], - [$type_enum, "\\\\fI\$1\\\\fP"], - [$type_struct, "\\\\fI\$1\\\\fP"], - [$type_typedef, "\\\\fI\$1\\\\fP"], - [$type_union, "\\\\fI\$1\\\\fP"], - [$type_param, "\\\\fI\$1\\\\fP"], - [$type_param_ref, "\\\\fI\$1\$2\\\\fP"], - [$type_member, "\\\\fI\$1\$2\$3\\\\fP"], - [$type_fallback, "\\\\fI\$1\\\\fP"] - ); -my $blankline_man = ""; - -# rst-mode -my @highlights_rst = ( - [$type_constant, "``\$1``"], - [$type_constant2, "``\$1``"], - # Note: need to escape () to avoid func matching later - [$type_member_func, "\\:c\\:type\\:`\$1\$2\$3\\\\(\\\\) <\$1>`"], - [$type_member, "\\:c\\:type\\:`\$1\$2\$3 <\$1>`"], - [$type_fp_param, "**\$1\\\\(\\\\)**"], - [$type_fp_param2, "**\$1\\\\(\\\\)**"], - [$type_func, "\$1()"], - [$type_enum, "\\:c\\:type\\:`\$1 <\$2>`"], - [$type_struct, "\\:c\\:type\\:`\$1 <\$2>`"], - [$type_typedef, "\\:c\\:type\\:`\$1 <\$2>`"], - [$type_union, "\\:c\\:type\\:`\$1 <\$2>`"], - # in rst this can refer to any type - [$type_fallback, "\\:c\\:type\\:`\$1`"], - [$type_param_ref, "**\$1\$2**"] - ); -my $blankline_rst = "\n"; - -# read arguments -if ($#ARGV == -1) { - usage(); -} - -my $kernelversion; -my ($sphinx_major, $sphinx_minor, $sphinx_patch); - -my $dohighlight = ""; - -my $verbose = 0; -my $Werror = 0; -my $output_mode = "rst"; -my $output_preformatted = 0; -my $no_doc_sections = 0; -my $enable_lineno = 0; -my @highlights = @highlights_rst; -my $blankline = $blankline_rst; -my $modulename = "Kernel API"; - -use constant { - OUTPUT_ALL => 0, # output all symbols and doc sections - OUTPUT_INCLUDE => 1, # output only specified symbols - OUTPUT_EXPORTED => 2, # output exported symbols - OUTPUT_INTERNAL => 3, # output non-exported symbols -}; -my $output_selection = OUTPUT_ALL; -my $show_not_found = 0; # No longer used - -my @export_file_list; - -my @build_time; -if (defined($ENV{'KBUILD_BUILD_TIMESTAMP'}) && - (my $seconds = `date -d"${ENV{'KBUILD_BUILD_TIMESTAMP'}}" +%s`) ne '') { - @build_time = gmtime($seconds); -} else { - @build_time = localtime; -} - -my $man_date = ('January', 'February', 'March', 'April', 'May', 'June', - 'July', 'August', 'September', 'October', - 'November', 'December')[$build_time[4]] . - " " . ($build_time[5]+1900); - -# Essentially these are globals. -# They probably want to be tidied up, made more localised or something. -# CAVEAT EMPTOR! Some of the others I localised may not want to be, which -# could cause "use of undefined value" or other bugs. -my ($function, %function_table, %parametertypes, $declaration_purpose); -my %nosymbol_table = (); -my $declaration_start_line; -my ($type, $declaration_name, $return_type); -my ($newsection, $newcontents, $prototype, $brcount, %source_map); - -if (defined($ENV{'KBUILD_VERBOSE'})) { - $verbose = "$ENV{'KBUILD_VERBOSE'}"; -} - -if (defined($ENV{'KDOC_WERROR'})) { - $Werror = "$ENV{'KDOC_WERROR'}"; -} - -if (defined($ENV{'KCFLAGS'})) { - my $kcflags = "$ENV{'KCFLAGS'}"; - - if ($kcflags =~ /Werror/) { - $Werror = 1; - } -} - -# Generated docbook code is inserted in a template at a point where -# docbook v3.1 requires a non-zero sequence of RefEntry's; see: -# https://www.oasis-open.org/docbook/documentation/reference/html/refentry.html -# We keep track of number of generated entries and generate a dummy -# if needs be to ensure the expanded template can be postprocessed -# into html. -my $section_counter = 0; - -my $lineprefix=""; - -# Parser states -use constant { - STATE_NORMAL => 0, # normal code - STATE_NAME => 1, # looking for function name - STATE_BODY_MAYBE => 2, # body - or maybe more description - STATE_BODY => 3, # the body of the comment - STATE_BODY_WITH_BLANK_LINE => 4, # the body, which has a blank line - STATE_PROTO => 5, # scanning prototype - STATE_DOCBLOCK => 6, # documentation block - STATE_INLINE => 7, # gathering doc outside main block -}; -my $state; -my $in_doc_sect; -my $leading_space; - -# Inline documentation state -use constant { - STATE_INLINE_NA => 0, # not applicable ($state != STATE_INLINE) - STATE_INLINE_NAME => 1, # looking for member name (@foo:) - STATE_INLINE_TEXT => 2, # looking for member documentation - STATE_INLINE_END => 3, # done - STATE_INLINE_ERROR => 4, # error - Comment without header was found. - # Spit a warning as it's not - # proper kernel-doc and ignore the rest. -}; -my $inline_doc_state; - -#declaration types: can be -# 'function', 'struct', 'union', 'enum', 'typedef' -my $decl_type; - -my $doc_start = '^/\*\*\s*$'; # Allow whitespace at end of comment start. -my $doc_end = '\*/'; -my $doc_com = '\s*\*\s*'; -my $doc_com_body = '\s*\* ?'; -my $doc_decl = $doc_com . '(\w+)'; -# @params and a strictly limited set of supported section names -my $doc_sect = $doc_com . - '\s*(\@[.\w]+|\@\.\.\.|description|context|returns?|notes?|examples?)\s*:(.*)'; -my $doc_content = $doc_com_body . '(.*)'; -my $doc_block = $doc_com . 'DOC:\s*(.*)?'; -my $doc_inline_start = '^\s*/\*\*\s*$'; -my $doc_inline_sect = '\s*\*\s*(@\s*[\w][\w\.]*\s*):(.*)'; -my $doc_inline_end = '^\s*\*/\s*$'; -my $doc_inline_oneline = '^\s*/\*\*\s*(@[\w\s]+):\s*(.*)\s*\*/\s*$'; -my $export_symbol = '^\s*EXPORT_SYMBOL(_GPL)?\s*\(\s*(\w+)\s*\)\s*;'; - -my %parameterdescs; -my %parameterdesc_start_lines; -my @parameterlist; -my %sections; -my @sectionlist; -my %section_start_lines; -my $sectcheck; -my $struct_actual; - -my $contents = ""; -my $new_start_line = 0; - -# the canonical section names. see also $doc_sect above. -my $section_default = "Description"; # default section -my $section_intro = "Introduction"; -my $section = $section_default; -my $section_context = "Context"; -my $section_return = "Return"; - -my $undescribed = "-- undescribed --"; - -reset_state(); - -while ($ARGV[0] =~ m/^--?(.*)/) { - my $cmd = $1; - shift @ARGV; - if ($cmd eq "man") { - $output_mode = "man"; - @highlights = @highlights_man; - $blankline = $blankline_man; - } elsif ($cmd eq "rst") { - $output_mode = "rst"; - @highlights = @highlights_rst; - $blankline = $blankline_rst; - } elsif ($cmd eq "none") { - $output_mode = "none"; - } elsif ($cmd eq "module") { # not needed for XML, inherits from calling document - $modulename = shift @ARGV; - } elsif ($cmd eq "function") { # to only output specific functions - $output_selection = OUTPUT_INCLUDE; - $function = shift @ARGV; - $function_table{$function} = 1; - } elsif ($cmd eq "nosymbol") { # Exclude specific symbols - my $symbol = shift @ARGV; - $nosymbol_table{$symbol} = 1; - } elsif ($cmd eq "export") { # only exported symbols - $output_selection = OUTPUT_EXPORTED; - %function_table = (); - } elsif ($cmd eq "internal") { # only non-exported symbols - $output_selection = OUTPUT_INTERNAL; - %function_table = (); - } elsif ($cmd eq "export-file") { - my $file = shift @ARGV; - push(@export_file_list, $file); - } elsif ($cmd eq "v") { - $verbose = 1; - } elsif ($cmd eq "Werror") { - $Werror = 1; - } elsif (($cmd eq "h") || ($cmd eq "help")) { - usage(); - } elsif ($cmd eq 'no-doc-sections') { - $no_doc_sections = 1; - } elsif ($cmd eq 'enable-lineno') { - $enable_lineno = 1; - } elsif ($cmd eq 'show-not-found') { - $show_not_found = 1; # A no-op but don't fail - } elsif ($cmd eq "sphinx-version") { - my $ver_string = shift @ARGV; - if ($ver_string =~ m/^(\d+)(\.\d+)?(\.\d+)?/) { - $sphinx_major = $1; - if (defined($2)) { - $sphinx_minor = substr($2,1); - } else { - $sphinx_minor = 0; - } - if (defined($3)) { - $sphinx_patch = substr($3,1) - } else { - $sphinx_patch = 0; - } - } else { - die "Sphinx version should either major.minor or major.minor.patch format\n"; - } - } else { - # Unknown argument - usage(); - } -} - -# continue execution near EOF; - -# The C domain dialect changed on Sphinx 3. So, we need to check the -# version in order to produce the right tags. -sub findprog($) -{ - foreach(split(/:/, $ENV{PATH})) { - return "$_/$_[0]" if(-x "$_/$_[0]"); - } -} - -sub get_sphinx_version() -{ - my $ver; - - my $cmd = "sphinx-build"; - if (!findprog($cmd)) { - my $cmd = "sphinx-build3"; - if (!findprog($cmd)) { - $sphinx_major = 1; - $sphinx_minor = 2; - $sphinx_patch = 0; - printf STDERR "Warning: Sphinx version not found. Using default (Sphinx version %d.%d.%d)\n", - $sphinx_major, $sphinx_minor, $sphinx_patch; - return; - } - } - - open IN, "$cmd --version 2>&1 |"; - while () { - if (m/^\s*sphinx-build\s+([\d]+)\.([\d\.]+)(\+\/[\da-f]+)?$/) { - $sphinx_major = $1; - $sphinx_minor = $2; - $sphinx_patch = $3; - last; - } - # Sphinx 1.2.x uses a different format - if (m/^\s*Sphinx.*\s+([\d]+)\.([\d\.]+)$/) { - $sphinx_major = $1; - $sphinx_minor = $2; - $sphinx_patch = $3; - last; - } - } - close IN; -} - -# get kernel version from env -sub get_kernel_version() { - my $version = 'unknown kernel version'; - - if (defined($ENV{'KERNELVERSION'})) { - $version = $ENV{'KERNELVERSION'}; - } - return $version; -} - -# -sub print_lineno { - my $lineno = shift; - if ($enable_lineno && defined($lineno)) { - print "#define LINENO " . $lineno . "\n"; - } -} -## -# dumps section contents to arrays/hashes intended for that purpose. -# -sub dump_section { - my $file = shift; - my $name = shift; - my $contents = join "\n", @_; - - if ($name =~ m/$type_param/) { - $name = $1; - $parameterdescs{$name} = $contents; - $sectcheck = $sectcheck . $name . " "; - $parameterdesc_start_lines{$name} = $new_start_line; - $new_start_line = 0; - } elsif ($name eq "@\.\.\.") { - $name = "..."; - $parameterdescs{$name} = $contents; - $sectcheck = $sectcheck . $name . " "; - $parameterdesc_start_lines{$name} = $new_start_line; - $new_start_line = 0; - } else { - if (defined($sections{$name}) && ($sections{$name} ne "")) { - # Only warn on user specified duplicate section names. - if ($name ne $section_default) { - print STDERR "${file}:$.: warning: duplicate section name '$name'\n"; - ++$warnings; - } - $sections{$name} .= $contents; - } else { - $sections{$name} = $contents; - push @sectionlist, $name; - $section_start_lines{$name} = $new_start_line; - $new_start_line = 0; - } - } -} - -## -# dump DOC: section after checking that it should go out -# -sub dump_doc_section { - my $file = shift; - my $name = shift; - my $contents = join "\n", @_; - - if ($no_doc_sections) { - return; - } - - return if (defined($nosymbol_table{$name})); - - if (($output_selection == OUTPUT_ALL) || - (($output_selection == OUTPUT_INCLUDE) && - defined($function_table{$name}))) - { - dump_section($file, $name, $contents); - output_blockhead({'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'module' => $modulename, - 'content-only' => ($output_selection != OUTPUT_ALL), }); - } -} - -## -# output function -# -# parameterdescs, a hash. -# function => "function name" -# parameterlist => @list of parameters -# parameterdescs => %parameter descriptions -# sectionlist => @list of sections -# sections => %section descriptions -# - -sub output_highlight { - my $contents = join "\n",@_; - my $line; - -# DEBUG -# if (!defined $contents) { -# use Carp; -# confess "output_highlight got called with no args?\n"; -# } - -# print STDERR "contents b4:$contents\n"; - eval $dohighlight; - die $@ if $@; -# print STDERR "contents af:$contents\n"; - - foreach $line (split "\n", $contents) { - if (! $output_preformatted) { - $line =~ s/^\s*//; - } - if ($line eq ""){ - if (! $output_preformatted) { - print $lineprefix, $blankline; - } - } else { - if ($output_mode eq "man" && substr($line, 0, 1) eq ".") { - print "\\&$line"; - } else { - print $lineprefix, $line; - } - } - print "\n"; - } -} - -## -# output function in man -sub output_function_man(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - my $count; - - print ".TH \"$args{'function'}\" 9 \"$args{'function'}\" \"$man_date\" \"Kernel Hacker's Manual\" LINUX\n"; - - print ".SH NAME\n"; - print $args{'function'} . " \\- " . $args{'purpose'} . "\n"; - - print ".SH SYNOPSIS\n"; - if ($args{'functiontype'} ne "") { - print ".B \"" . $args{'functiontype'} . "\" " . $args{'function'} . "\n"; - } else { - print ".B \"" . $args{'function'} . "\n"; - } - $count = 0; - my $parenth = "("; - my $post = ","; - foreach my $parameter (@{$args{'parameterlist'}}) { - if ($count == $#{$args{'parameterlist'}}) { - $post = ");"; - } - $type = $args{'parametertypes'}{$parameter}; - if ($type =~ m/([^\(]*\(\*)\s*\)\s*\(([^\)]*)\)/) { - # pointer-to-function - print ".BI \"" . $parenth . $1 . "\" " . " \") (" . $2 . ")" . $post . "\"\n"; - } else { - $type =~ s/([^\*])$/$1 /; - print ".BI \"" . $parenth . $type . "\" " . " \"" . $post . "\"\n"; - } - $count++; - $parenth = ""; - } - - print ".SH ARGUMENTS\n"; - foreach $parameter (@{$args{'parameterlist'}}) { - my $parameter_name = $parameter; - $parameter_name =~ s/\[.*//; - - print ".IP \"" . $parameter . "\" 12\n"; - output_highlight($args{'parameterdescs'}{$parameter_name}); - } - foreach $section (@{$args{'sectionlist'}}) { - print ".SH \"", uc $section, "\"\n"; - output_highlight($args{'sections'}{$section}); - } -} - -## -# output enum in man -sub output_enum_man(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - my $count; - - print ".TH \"$args{'module'}\" 9 \"enum $args{'enum'}\" \"$man_date\" \"API Manual\" LINUX\n"; - - print ".SH NAME\n"; - print "enum " . $args{'enum'} . " \\- " . $args{'purpose'} . "\n"; - - print ".SH SYNOPSIS\n"; - print "enum " . $args{'enum'} . " {\n"; - $count = 0; - foreach my $parameter (@{$args{'parameterlist'}}) { - print ".br\n.BI \" $parameter\"\n"; - if ($count == $#{$args{'parameterlist'}}) { - print "\n};\n"; - last; - } - else { - print ", \n.br\n"; - } - $count++; - } - - print ".SH Constants\n"; - foreach $parameter (@{$args{'parameterlist'}}) { - my $parameter_name = $parameter; - $parameter_name =~ s/\[.*//; - - print ".IP \"" . $parameter . "\" 12\n"; - output_highlight($args{'parameterdescs'}{$parameter_name}); - } - foreach $section (@{$args{'sectionlist'}}) { - print ".SH \"$section\"\n"; - output_highlight($args{'sections'}{$section}); - } -} - -## -# output struct in man -sub output_struct_man(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - - print ".TH \"$args{'module'}\" 9 \"" . $args{'type'} . " " . $args{'struct'} . "\" \"$man_date\" \"API Manual\" LINUX\n"; - - print ".SH NAME\n"; - print $args{'type'} . " " . $args{'struct'} . " \\- " . $args{'purpose'} . "\n"; - - my $declaration = $args{'definition'}; - $declaration =~ s/\t/ /g; - $declaration =~ s/\n/"\n.br\n.BI \"/g; - print ".SH SYNOPSIS\n"; - print $args{'type'} . " " . $args{'struct'} . " {\n.br\n"; - print ".BI \"$declaration\n};\n.br\n\n"; - - print ".SH Members\n"; - foreach $parameter (@{$args{'parameterlist'}}) { - ($parameter =~ /^#/) && next; - - my $parameter_name = $parameter; - $parameter_name =~ s/\[.*//; - - ($args{'parameterdescs'}{$parameter_name} ne $undescribed) || next; - print ".IP \"" . $parameter . "\" 12\n"; - output_highlight($args{'parameterdescs'}{$parameter_name}); - } - foreach $section (@{$args{'sectionlist'}}) { - print ".SH \"$section\"\n"; - output_highlight($args{'sections'}{$section}); - } -} - -## -# output typedef in man -sub output_typedef_man(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - - print ".TH \"$args{'module'}\" 9 \"$args{'typedef'}\" \"$man_date\" \"API Manual\" LINUX\n"; - - print ".SH NAME\n"; - print "typedef " . $args{'typedef'} . " \\- " . $args{'purpose'} . "\n"; - - foreach $section (@{$args{'sectionlist'}}) { - print ".SH \"$section\"\n"; - output_highlight($args{'sections'}{$section}); - } -} - -sub output_blockhead_man(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - my $count; - - print ".TH \"$args{'module'}\" 9 \"$args{'module'}\" \"$man_date\" \"API Manual\" LINUX\n"; - - foreach $section (@{$args{'sectionlist'}}) { - print ".SH \"$section\"\n"; - output_highlight($args{'sections'}{$section}); - } -} - -## -# output in restructured text -# - -# -# This could use some work; it's used to output the DOC: sections, and -# starts by putting out the name of the doc section itself, but that tends -# to duplicate a header already in the template file. -# -sub output_blockhead_rst(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - - foreach $section (@{$args{'sectionlist'}}) { - next if (defined($nosymbol_table{$section})); - - if ($output_selection != OUTPUT_INCLUDE) { - print "**$section**\n\n"; - } - print_lineno($section_start_lines{$section}); - output_highlight_rst($args{'sections'}{$section}); - print "\n"; - } -} - -# -# Apply the RST highlights to a sub-block of text. -# -sub highlight_block($) { - # The dohighlight kludge requires the text be called $contents - my $contents = shift; - eval $dohighlight; - die $@ if $@; - return $contents; -} - -# -# Regexes used only here. -# -my $sphinx_literal = '^[^.].*::$'; -my $sphinx_cblock = '^\.\.\ +code-block::'; - -sub output_highlight_rst { - my $input = join "\n",@_; - my $output = ""; - my $line; - my $in_literal = 0; - my $litprefix; - my $block = ""; - - foreach $line (split "\n",$input) { - # - # If we're in a literal block, see if we should drop out - # of it. Otherwise pass the line straight through unmunged. - # - if ($in_literal) { - if (! ($line =~ /^\s*$/)) { - # - # If this is the first non-blank line in a literal - # block we need to figure out what the proper indent is. - # - if ($litprefix eq "") { - $line =~ /^(\s*)/; - $litprefix = '^' . $1; - $output .= $line . "\n"; - } elsif (! ($line =~ /$litprefix/)) { - $in_literal = 0; - } else { - $output .= $line . "\n"; - } - } else { - $output .= $line . "\n"; - } - } - # - # Not in a literal block (or just dropped out) - # - if (! $in_literal) { - $block .= $line . "\n"; - if (($line =~ /$sphinx_literal/) || ($line =~ /$sphinx_cblock/)) { - $in_literal = 1; - $litprefix = ""; - $output .= highlight_block($block); - $block = "" - } - } - } - - if ($block) { - $output .= highlight_block($block); - } - foreach $line (split "\n", $output) { - print $lineprefix . $line . "\n"; - } -} - -sub output_function_rst(%) { - my %args = %{$_[0]}; - my ($parameter, $section); - my $oldprefix = $lineprefix; - my $start = ""; - my $is_macro = 0; - - if ($sphinx_major < 3) { - if ($args{'typedef'}) { - print ".. c:type:: ". $args{'function'} . "\n\n"; - print_lineno($declaration_start_line); - print " **Typedef**: "; - $lineprefix = ""; - output_highlight_rst($args{'purpose'}); - $start = "\n\n**Syntax**\n\n ``"; - $is_macro = 1; - } else { - print ".. c:function:: "; - } - } else { - if ($args{'typedef'} || $args{'functiontype'} eq "") { - $is_macro = 1; - print ".. c:macro:: ". $args{'function'} . "\n\n"; - } else { - print ".. c:function:: "; - } - - if ($args{'typedef'}) { - print_lineno($declaration_start_line); - print " **Typedef**: "; - $lineprefix = ""; - output_highlight_rst($args{'purpose'}); - $start = "\n\n**Syntax**\n\n ``"; - } else { - print "``" if ($is_macro); - } - } - if ($args{'functiontype'} ne "") { - $start .= $args{'functiontype'} . " " . $args{'function'} . " ("; - } else { - $start .= $args{'function'} . " ("; - } - print $start; - - my $count = 0; - foreach my $parameter (@{$args{'parameterlist'}}) { - if ($count ne 0) { - print ", "; - } - $count++; - $type = $args{'parametertypes'}{$parameter}; - - if ($type =~ m/([^\(]*\(\*)\s*\)\s*\(([^\)]*)\)/) { - # pointer-to-function - print $1 . $parameter . ") (" . $2 . ")"; - } else { - print $type; - } - } - if ($is_macro) { - print ")``\n\n"; - } else { - print ")\n\n"; - } - if (!$args{'typedef'}) { - print_lineno($declaration_start_line); - $lineprefix = " "; - output_highlight_rst($args{'purpose'}); - print "\n"; - } - - print "**Parameters**\n\n"; - $lineprefix = " "; - foreach $parameter (@{$args{'parameterlist'}}) { - my $parameter_name = $parameter; - $parameter_name =~ s/\[.*//; - $type = $args{'parametertypes'}{$parameter}; - - if ($type ne "") { - print "``$type``\n"; - } else { - print "``$parameter``\n"; - } - - print_lineno($parameterdesc_start_lines{$parameter_name}); - - if (defined($args{'parameterdescs'}{$parameter_name}) && - $args{'parameterdescs'}{$parameter_name} ne $undescribed) { - output_highlight_rst($args{'parameterdescs'}{$parameter_name}); - } else { - print " *undescribed*\n"; - } - print "\n"; - } - - $lineprefix = $oldprefix; - output_section_rst(@_); -} - -sub output_section_rst(%) { - my %args = %{$_[0]}; - my $section; - my $oldprefix = $lineprefix; - $lineprefix = ""; - - foreach $section (@{$args{'sectionlist'}}) { - print "**$section**\n\n"; - print_lineno($section_start_lines{$section}); - output_highlight_rst($args{'sections'}{$section}); - print "\n"; - } - print "\n"; - $lineprefix = $oldprefix; -} - -sub output_enum_rst(%) { - my %args = %{$_[0]}; - my ($parameter); - my $oldprefix = $lineprefix; - my $count; - - if ($sphinx_major < 3) { - my $name = "enum " . $args{'enum'}; - print "\n\n.. c:type:: " . $name . "\n\n"; - } else { - my $name = $args{'enum'}; - print "\n\n.. c:enum:: " . $name . "\n\n"; - } - print_lineno($declaration_start_line); - $lineprefix = " "; - output_highlight_rst($args{'purpose'}); - print "\n"; - - print "**Constants**\n\n"; - $lineprefix = " "; - foreach $parameter (@{$args{'parameterlist'}}) { - print "``$parameter``\n"; - if ($args{'parameterdescs'}{$parameter} ne $undescribed) { - output_highlight_rst($args{'parameterdescs'}{$parameter}); - } else { - print " *undescribed*\n"; - } - print "\n"; - } - - $lineprefix = $oldprefix; - output_section_rst(@_); -} - -sub output_typedef_rst(%) { - my %args = %{$_[0]}; - my ($parameter); - my $oldprefix = $lineprefix; - my $name; - - if ($sphinx_major < 3) { - $name = "typedef " . $args{'typedef'}; - } else { - $name = $args{'typedef'}; - } - print "\n\n.. c:type:: " . $name . "\n\n"; - print_lineno($declaration_start_line); - $lineprefix = " "; - output_highlight_rst($args{'purpose'}); - print "\n"; - - $lineprefix = $oldprefix; - output_section_rst(@_); -} - -sub output_struct_rst(%) { - my %args = %{$_[0]}; - my ($parameter); - my $oldprefix = $lineprefix; - - if ($sphinx_major < 3) { - my $name = $args{'type'} . " " . $args{'struct'}; - print "\n\n.. c:type:: " . $name . "\n\n"; - } else { - my $name = $args{'struct'}; - if ($args{'type'} eq 'union') { - print "\n\n.. c:union:: " . $name . "\n\n"; - } else { - print "\n\n.. c:struct:: " . $name . "\n\n"; - } - } - print_lineno($declaration_start_line); - $lineprefix = " "; - output_highlight_rst($args{'purpose'}); - print "\n"; - - print "**Definition**\n\n"; - print "::\n\n"; - my $declaration = $args{'definition'}; - $declaration =~ s/\t/ /g; - print " " . $args{'type'} . " " . $args{'struct'} . " {\n$declaration };\n\n"; - - print "**Members**\n\n"; - $lineprefix = " "; - foreach $parameter (@{$args{'parameterlist'}}) { - ($parameter =~ /^#/) && next; - - my $parameter_name = $parameter; - $parameter_name =~ s/\[.*//; - - ($args{'parameterdescs'}{$parameter_name} ne $undescribed) || next; - $type = $args{'parametertypes'}{$parameter}; - print_lineno($parameterdesc_start_lines{$parameter_name}); - print "``" . $parameter . "``\n"; - output_highlight_rst($args{'parameterdescs'}{$parameter_name}); - print "\n"; - } - print "\n"; - - $lineprefix = $oldprefix; - output_section_rst(@_); -} - -## none mode output functions - -sub output_function_none(%) { -} - -sub output_enum_none(%) { -} - -sub output_typedef_none(%) { -} - -sub output_struct_none(%) { -} - -sub output_blockhead_none(%) { -} - -## -# generic output function for all types (function, struct/union, typedef, enum); -# calls the generated, variable output_ function name based on -# functype and output_mode -sub output_declaration { - no strict 'refs'; - my $name = shift; - my $functype = shift; - my $func = "output_${functype}_$output_mode"; - - return if (defined($nosymbol_table{$name})); - - if (($output_selection == OUTPUT_ALL) || - (($output_selection == OUTPUT_INCLUDE || - $output_selection == OUTPUT_EXPORTED) && - defined($function_table{$name})) || - ($output_selection == OUTPUT_INTERNAL && - !($functype eq "function" && defined($function_table{$name})))) - { - &$func(@_); - $section_counter++; - } -} - -## -# generic output function - calls the right one based on current output mode. -sub output_blockhead { - no strict 'refs'; - my $func = "output_blockhead_" . $output_mode; - &$func(@_); - $section_counter++; -} - -## -# takes a declaration (struct, union, enum, typedef) and -# invokes the right handler. NOT called for functions. -sub dump_declaration($$) { - no strict 'refs'; - my ($prototype, $file) = @_; - my $func = "dump_" . $decl_type; - &$func(@_); -} - -sub dump_union($$) { - dump_struct(@_); -} - -sub dump_struct($$) { - my $x = shift; - my $file = shift; - - if ($x =~ /(struct|union)\s+(\w+)\s*\{(.*)\}(\s*(__packed|__aligned|____cacheline_aligned_in_smp|____cacheline_aligned|__attribute__\s*\(\([a-z0-9,_\s\(\)]*\)\)))*/) { - my $decl_type = $1; - $declaration_name = $2; - my $members = $3; - - # ignore members marked private: - $members =~ s/\/\*\s*private:.*?\/\*\s*public:.*?\*\///gosi; - $members =~ s/\/\*\s*private:.*//gosi; - # strip comments: - $members =~ s/\/\*.*?\*\///gos; - # strip attributes - $members =~ s/\s*__attribute__\s*\(\([a-z0-9,_\*\s\(\)]*\)\)/ /gi; - $members =~ s/\s*__aligned\s*\([^;]*\)/ /gos; - $members =~ s/\s*__packed\s*/ /gos; - $members =~ s/\s*CRYPTO_MINALIGN_ATTR/ /gos; - $members =~ s/\s*____cacheline_aligned_in_smp/ /gos; - $members =~ s/\s*____cacheline_aligned/ /gos; - - # replace DECLARE_BITMAP - $members =~ s/__ETHTOOL_DECLARE_LINK_MODE_MASK\s*\(([^\)]+)\)/DECLARE_BITMAP($1, __ETHTOOL_LINK_MODE_MASK_NBITS)/gos; - $members =~ s/DECLARE_BITMAP\s*\(([^,)]+),\s*([^,)]+)\)/unsigned long $1\[BITS_TO_LONGS($2)\]/gos; - # replace DECLARE_HASHTABLE - $members =~ s/DECLARE_HASHTABLE\s*\(([^,)]+),\s*([^,)]+)\)/unsigned long $1\[1 << (($2) - 1)\]/gos; - # replace DECLARE_KFIFO - $members =~ s/DECLARE_KFIFO\s*\(([^,)]+),\s*([^,)]+),\s*([^,)]+)\)/$2 \*$1/gos; - # replace DECLARE_KFIFO_PTR - $members =~ s/DECLARE_KFIFO_PTR\s*\(([^,)]+),\s*([^,)]+)\)/$2 \*$1/gos; - - my $declaration = $members; - - # Split nested struct/union elements as newer ones - while ($members =~ m/(struct|union)([^\{\};]+)\{([^\{\}]*)\}([^\{\}\;]*)\;/) { - my $newmember; - my $maintype = $1; - my $ids = $4; - my $content = $3; - foreach my $id(split /,/, $ids) { - $newmember .= "$maintype $id; "; - - $id =~ s/[:\[].*//; - $id =~ s/^\s*\**(\S+)\s*/$1/; - foreach my $arg (split /;/, $content) { - next if ($arg =~ m/^\s*$/); - if ($arg =~ m/^([^\(]+\(\*?\s*)([\w\.]*)(\s*\).*)/) { - # pointer-to-function - my $type = $1; - my $name = $2; - my $extra = $3; - next if (!$name); - if ($id =~ m/^\s*$/) { - # anonymous struct/union - $newmember .= "$type$name$extra; "; - } else { - $newmember .= "$type$id.$name$extra; "; - } - } else { - my $type; - my $names; - $arg =~ s/^\s+//; - $arg =~ s/\s+$//; - # Handle bitmaps - $arg =~ s/:\s*\d+\s*//g; - # Handle arrays - $arg =~ s/\[.*\]//g; - # The type may have multiple words, - # and multiple IDs can be defined, like: - # const struct foo, *bar, foobar - # So, we remove spaces when parsing the - # names, in order to match just names - # and commas for the names - $arg =~ s/\s*,\s*/,/g; - if ($arg =~ m/(.*)\s+([\S+,]+)/) { - $type = $1; - $names = $2; - } else { - $newmember .= "$arg; "; - next; - } - foreach my $name (split /,/, $names) { - $name =~ s/^\s*\**(\S+)\s*/$1/; - next if (($name =~ m/^\s*$/)); - if ($id =~ m/^\s*$/) { - # anonymous struct/union - $newmember .= "$type $name; "; - } else { - $newmember .= "$type $id.$name; "; - } - } - } - } - } - $members =~ s/(struct|union)([^\{\};]+)\{([^\{\}]*)\}([^\{\}\;]*)\;/$newmember/; - } - - # Ignore other nested elements, like enums - $members =~ s/(\{[^\{\}]*\})//g; - - create_parameterlist($members, ';', $file, $declaration_name); - check_sections($file, $declaration_name, $decl_type, $sectcheck, $struct_actual); - - # Adjust declaration for better display - $declaration =~ s/([\{;])/$1\n/g; - $declaration =~ s/\}\s+;/};/g; - # Better handle inlined enums - do {} while ($declaration =~ s/(enum\s+\{[^\}]+),([^\n])/$1,\n$2/); - - my @def_args = split /\n/, $declaration; - my $level = 1; - $declaration = ""; - foreach my $clause (@def_args) { - $clause =~ s/^\s+//; - $clause =~ s/\s+$//; - $clause =~ s/\s+/ /; - next if (!$clause); - $level-- if ($clause =~ m/(\})/ && $level > 1); - if (!($clause =~ m/^\s*#/)) { - $declaration .= "\t" x $level; - } - $declaration .= "\t" . $clause . "\n"; - $level++ if ($clause =~ m/(\{)/ && !($clause =~m/\}/)); - } - output_declaration($declaration_name, - 'struct', - {'struct' => $declaration_name, - 'module' => $modulename, - 'definition' => $declaration, - 'parameterlist' => \@parameterlist, - 'parameterdescs' => \%parameterdescs, - 'parametertypes' => \%parametertypes, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose, - 'type' => $decl_type - }); - } - else { - print STDERR "${file}:$.: error: Cannot parse struct or union!\n"; - ++$errors; - } -} - - -sub show_warnings($$) { - my $functype = shift; - my $name = shift; - - return 0 if (defined($nosymbol_table{$name})); - - return 1 if ($output_selection == OUTPUT_ALL); - - if ($output_selection == OUTPUT_EXPORTED) { - if (defined($function_table{$name})) { - return 1; - } else { - return 0; - } - } - if ($output_selection == OUTPUT_INTERNAL) { - if (!($functype eq "function" && defined($function_table{$name}))) { - return 1; - } else { - return 0; - } - } - if ($output_selection == OUTPUT_INCLUDE) { - if (defined($function_table{$name})) { - return 1; - } else { - return 0; - } - } - die("Please add the new output type at show_warnings()"); -} - -sub dump_enum($$) { - my $x = shift; - my $file = shift; - my $members; - - - $x =~ s@/\*.*?\*/@@gos; # strip comments. - # strip #define macros inside enums - $x =~ s@#\s*((define|ifdef)\s+|endif)[^;]*;@@gos; - - if ($x =~ /typedef\s+enum\s*\{(.*)\}\s*(\w*)\s*;/) { - $declaration_name = $2; - $members = $1; - } elsif ($x =~ /enum\s+(\w*)\s*\{(.*)\}/) { - $declaration_name = $1; - $members = $2; - } - - if ($declaration_name) { - my %_members; - - $members =~ s/\s+$//; - - foreach my $arg (split ',', $members) { - $arg =~ s/^\s*(\w+).*/$1/; - push @parameterlist, $arg; - if (!$parameterdescs{$arg}) { - $parameterdescs{$arg} = $undescribed; - if (show_warnings("enum", $declaration_name)) { - print STDERR "${file}:$.: warning: Enum value '$arg' not described in enum '$declaration_name'\n"; - } - } - $_members{$arg} = 1; - } - - while (my ($k, $v) = each %parameterdescs) { - if (!exists($_members{$k})) { - if (show_warnings("enum", $declaration_name)) { - print STDERR "${file}:$.: warning: Excess enum value '$k' description in '$declaration_name'\n"; - } - } - } - - output_declaration($declaration_name, - 'enum', - {'enum' => $declaration_name, - 'module' => $modulename, - 'parameterlist' => \@parameterlist, - 'parameterdescs' => \%parameterdescs, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose - }); - } else { - print STDERR "${file}:$.: error: Cannot parse enum!\n"; - ++$errors; - } -} - -my $typedef_type = qr { ((?:\s+[\w\*]+){1,8})\s* }x; -my $typedef_ident = qr { \*?\s*(\w\S+)\s* }x; -my $typedef_args = qr { \s*\((.*)\); }x; - -my $typedef1 = qr { typedef$typedef_type\($typedef_ident\)$typedef_args }x; -my $typedef2 = qr { typedef$typedef_type$typedef_ident$typedef_args }x; - -sub dump_typedef($$) { - my $x = shift; - my $file = shift; - - $x =~ s@/\*.*?\*/@@gos; # strip comments. - - # Parse function typedef prototypes - if ($x =~ $typedef1 || $x =~ $typedef2) { - $return_type = $1; - $declaration_name = $2; - my $args = $3; - $return_type =~ s/^\s+//; - - create_parameterlist($args, ',', $file, $declaration_name); - - output_declaration($declaration_name, - 'function', - {'function' => $declaration_name, - 'typedef' => 1, - 'module' => $modulename, - 'functiontype' => $return_type, - 'parameterlist' => \@parameterlist, - 'parameterdescs' => \%parameterdescs, - 'parametertypes' => \%parametertypes, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose - }); - return; - } - - while (($x =~ /\(*.\)\s*;$/) || ($x =~ /\[*.\]\s*;$/)) { - $x =~ s/\(*.\)\s*;$/;/; - $x =~ s/\[*.\]\s*;$/;/; - } - - if ($x =~ /typedef.*\s+(\w+)\s*;/) { - $declaration_name = $1; - - output_declaration($declaration_name, - 'typedef', - {'typedef' => $declaration_name, - 'module' => $modulename, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose - }); - } - else { - print STDERR "${file}:$.: error: Cannot parse typedef!\n"; - ++$errors; - } -} - -sub save_struct_actual($) { - my $actual = shift; - - # strip all spaces from the actual param so that it looks like one string item - $actual =~ s/\s*//g; - $struct_actual = $struct_actual . $actual . " "; -} - -sub create_parameterlist($$$$) { - my $args = shift; - my $splitter = shift; - my $file = shift; - my $declaration_name = shift; - my $type; - my $param; - - # temporarily replace commas inside function pointer definition - while ($args =~ /(\([^\),]+),/) { - $args =~ s/(\([^\),]+),/$1#/g; - } - - foreach my $arg (split($splitter, $args)) { - # strip comments - $arg =~ s/\/\*.*\*\///; - # strip leading/trailing spaces - $arg =~ s/^\s*//; - $arg =~ s/\s*$//; - $arg =~ s/\s+/ /; - - if ($arg =~ /^#/) { - # Treat preprocessor directive as a typeless variable just to fill - # corresponding data structures "correctly". Catch it later in - # output_* subs. - push_parameter($arg, "", "", $file); - } elsif ($arg =~ m/\(.+\)\s*\(/) { - # pointer-to-function - $arg =~ tr/#/,/; - $arg =~ m/[^\(]+\(\*?\s*([\w\.]*)\s*\)/; - $param = $1; - $type = $arg; - $type =~ s/([^\(]+\(\*?)\s*$param/$1/; - save_struct_actual($param); - push_parameter($param, $type, $arg, $file, $declaration_name); - } elsif ($arg) { - $arg =~ s/\s*:\s*/:/g; - $arg =~ s/\s*\[/\[/g; - - my @args = split('\s*,\s*', $arg); - if ($args[0] =~ m/\*/) { - $args[0] =~ s/(\*+)\s*/ $1/; - } - - my @first_arg; - if ($args[0] =~ /^(.*\s+)(.*?\[.*\].*)$/) { - shift @args; - push(@first_arg, split('\s+', $1)); - push(@first_arg, $2); - } else { - @first_arg = split('\s+', shift @args); - } - - unshift(@args, pop @first_arg); - $type = join " ", @first_arg; - - foreach $param (@args) { - if ($param =~ m/^(\*+)\s*(.*)/) { - save_struct_actual($2); - - push_parameter($2, "$type $1", $arg, $file, $declaration_name); - } - elsif ($param =~ m/(.*?):(\d+)/) { - if ($type ne "") { # skip unnamed bit-fields - save_struct_actual($1); - push_parameter($1, "$type:$2", $arg, $file, $declaration_name) - } - } - else { - save_struct_actual($param); - push_parameter($param, $type, $arg, $file, $declaration_name); - } - } - } - } -} - -sub push_parameter($$$$$) { - my $param = shift; - my $type = shift; - my $org_arg = shift; - my $file = shift; - my $declaration_name = shift; - - if (($anon_struct_union == 1) && ($type eq "") && - ($param eq "}")) { - return; # ignore the ending }; from anon. struct/union - } - - $anon_struct_union = 0; - $param =~ s/[\[\)].*//; - - if ($type eq "" && $param =~ /\.\.\.$/) - { - if ($param =~ /\w\.\.\.$/) { - # for named variable parameters of the form `x...`, remove the dots - $param =~ s/\.\.\.$//; - } else { - # handles unnamed variable parameters - $param = "..."; - } - if (!defined $parameterdescs{$param} || $parameterdescs{$param} eq "") { - $parameterdescs{$param} = "variable arguments"; - } - } - elsif ($type eq "" && ($param eq "" or $param eq "void")) - { - $param="void"; - $parameterdescs{void} = "no arguments"; - } - elsif ($type eq "" && ($param eq "struct" or $param eq "union")) - # handle unnamed (anonymous) union or struct: - { - $type = $param; - $param = "{unnamed_" . $param . "}"; - $parameterdescs{$param} = "anonymous\n"; - $anon_struct_union = 1; - } - - # warn if parameter has no description - # (but ignore ones starting with # as these are not parameters - # but inline preprocessor statements); - # Note: It will also ignore void params and unnamed structs/unions - if (!defined $parameterdescs{$param} && $param !~ /^#/) { - $parameterdescs{$param} = $undescribed; - - if (show_warnings($type, $declaration_name) && $param !~ /\./) { - print STDERR - "${file}:$.: warning: Function parameter or member '$param' not described in '$declaration_name'\n"; - ++$warnings; - } - } - - # strip spaces from $param so that it is one continuous string - # on @parameterlist; - # this fixes a problem where check_sections() cannot find - # a parameter like "addr[6 + 2]" because it actually appears - # as "addr[6", "+", "2]" on the parameter list; - # but it's better to maintain the param string unchanged for output, - # so just weaken the string compare in check_sections() to ignore - # "[blah" in a parameter string; - ###$param =~ s/\s*//g; - push @parameterlist, $param; - $org_arg =~ s/\s\s+/ /g; - $parametertypes{$param} = $org_arg; -} - -sub check_sections($$$$$) { - my ($file, $decl_name, $decl_type, $sectcheck, $prmscheck) = @_; - my @sects = split ' ', $sectcheck; - my @prms = split ' ', $prmscheck; - my $err; - my ($px, $sx); - my $prm_clean; # strip trailing "[array size]" and/or beginning "*" - - foreach $sx (0 .. $#sects) { - $err = 1; - foreach $px (0 .. $#prms) { - $prm_clean = $prms[$px]; - $prm_clean =~ s/\[.*\]//; - $prm_clean =~ s/__attribute__\s*\(\([a-z,_\*\s\(\)]*\)\)//i; - # ignore array size in a parameter string; - # however, the original param string may contain - # spaces, e.g.: addr[6 + 2] - # and this appears in @prms as "addr[6" since the - # parameter list is split at spaces; - # hence just ignore "[..." for the sections check; - $prm_clean =~ s/\[.*//; - - ##$prm_clean =~ s/^\**//; - if ($prm_clean eq $sects[$sx]) { - $err = 0; - last; - } - } - if ($err) { - if ($decl_type eq "function") { - print STDERR "${file}:$.: warning: " . - "Excess function parameter " . - "'$sects[$sx]' " . - "description in '$decl_name'\n"; - ++$warnings; - } - } - } -} - -## -# Checks the section describing the return value of a function. -sub check_return_section { - my $file = shift; - my $declaration_name = shift; - my $return_type = shift; - - # Ignore an empty return type (It's a macro) - # Ignore functions with a "void" return type. (But don't ignore "void *") - if (($return_type eq "") || ($return_type =~ /void\s*\w*\s*$/)) { - return; - } - - if (!defined($sections{$section_return}) || - $sections{$section_return} eq "") { - print STDERR "${file}:$.: warning: " . - "No description found for return value of " . - "'$declaration_name'\n"; - ++$warnings; - } -} - -## -# takes a function prototype and the name of the current file being -# processed and spits out all the details stored in the global -# arrays/hashes. -sub dump_function($$) { - my $prototype = shift; - my $file = shift; - my $noret = 0; - - print_lineno($new_start_line); - - $prototype =~ s/^static +//; - $prototype =~ s/^extern +//; - $prototype =~ s/^asmlinkage +//; - $prototype =~ s/^inline +//; - $prototype =~ s/^__inline__ +//; - $prototype =~ s/^__inline +//; - $prototype =~ s/^__always_inline +//; - $prototype =~ s/^noinline +//; - $prototype =~ s/__init +//; - $prototype =~ s/__init_or_module +//; - $prototype =~ s/__meminit +//; - $prototype =~ s/__must_check +//; - $prototype =~ s/__weak +//; - $prototype =~ s/__sched +//; - $prototype =~ s/__printf\s*\(\s*\d*\s*,\s*\d*\s*\) +//; - my $define = $prototype =~ s/^#\s*define\s+//; #ak added - $prototype =~ s/__attribute__\s*\(\( - (?: - [\w\s]++ # attribute name - (?:\([^)]*+\))? # attribute arguments - \s*+,? # optional comma at the end - )+ - \)\)\s+//x; - - # Strip QEMU specific compiler annotations - $prototype =~ s/QEMU_[A-Z_]+ +//; - - # Yes, this truly is vile. We are looking for: - # 1. Return type (may be nothing if we're looking at a macro) - # 2. Function name - # 3. Function parameters. - # - # All the while we have to watch out for function pointer parameters - # (which IIRC is what the two sections are for), C types (these - # regexps don't even start to express all the possibilities), and - # so on. - # - # If you mess with these regexps, it's a good idea to check that - # the following functions' documentation still comes out right: - # - parport_register_device (function pointer parameters) - # - atomic_set (macro) - # - pci_match_device, __copy_to_user (long return type) - - if ($define && $prototype =~ m/^()([a-zA-Z0-9_~:]+)\s+/) { - # This is an object-like macro, it has no return type and no parameter - # list. - # Function-like macros are not allowed to have spaces between - # declaration_name and opening parenthesis (notice the \s+). - $return_type = $1; - $declaration_name = $2; - $noret = 1; - } elsif ($prototype =~ m/^()([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+\s+\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\(]*)\)/ || - $prototype =~ m/^()([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+\s+\w+)\s+([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s+\w+\s+\w+\s*\*+)\s*([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/ || - $prototype =~ m/^(\w+\s+\w+\s*\*+\s*\w+\s*\*+\s*)\s*([a-zA-Z0-9_~:]+)\s*\(([^\{]*)\)/) { - $return_type = $1; - $declaration_name = $2; - my $args = $3; - - create_parameterlist($args, ',', $file, $declaration_name); - } else { - print STDERR "${file}:$.: warning: cannot understand function prototype: '$prototype'\n"; - return; - } - - my $prms = join " ", @parameterlist; - check_sections($file, $declaration_name, "function", $sectcheck, $prms); - - # This check emits a lot of warnings at the moment, because many - # functions don't have a 'Return' doc section. So until the number - # of warnings goes sufficiently down, the check is only performed in - # verbose mode. - # TODO: always perform the check. - if ($verbose && !$noret) { - check_return_section($file, $declaration_name, $return_type); - } - - # The function parser can be called with a typedef parameter. - # Handle it. - if ($return_type =~ /typedef/) { - output_declaration($declaration_name, - 'function', - {'function' => $declaration_name, - 'typedef' => 1, - 'module' => $modulename, - 'functiontype' => $return_type, - 'parameterlist' => \@parameterlist, - 'parameterdescs' => \%parameterdescs, - 'parametertypes' => \%parametertypes, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose - }); - } else { - output_declaration($declaration_name, - 'function', - {'function' => $declaration_name, - 'module' => $modulename, - 'functiontype' => $return_type, - 'parameterlist' => \@parameterlist, - 'parameterdescs' => \%parameterdescs, - 'parametertypes' => \%parametertypes, - 'sectionlist' => \@sectionlist, - 'sections' => \%sections, - 'purpose' => $declaration_purpose - }); - } -} - -sub reset_state { - $function = ""; - %parameterdescs = (); - %parametertypes = (); - @parameterlist = (); - %sections = (); - @sectionlist = (); - $sectcheck = ""; - $struct_actual = ""; - $prototype = ""; - - $state = STATE_NORMAL; - $inline_doc_state = STATE_INLINE_NA; -} - -sub tracepoint_munge($) { - my $file = shift; - my $tracepointname = 0; - my $tracepointargs = 0; - - if ($prototype =~ m/TRACE_EVENT\((.*?),/) { - $tracepointname = $1; - } - if ($prototype =~ m/DEFINE_SINGLE_EVENT\((.*?),/) { - $tracepointname = $1; - } - if ($prototype =~ m/DEFINE_EVENT\((.*?),(.*?),/) { - $tracepointname = $2; - } - $tracepointname =~ s/^\s+//; #strip leading whitespace - if ($prototype =~ m/TP_PROTO\((.*?)\)/) { - $tracepointargs = $1; - } - if (($tracepointname eq 0) || ($tracepointargs eq 0)) { - print STDERR "${file}:$.: warning: Unrecognized tracepoint format: \n". - "$prototype\n"; - } else { - $prototype = "static inline void trace_$tracepointname($tracepointargs)"; - } -} - -sub syscall_munge() { - my $void = 0; - - $prototype =~ s@[\r\n]+@ @gos; # strip newlines/CR's -## if ($prototype =~ m/SYSCALL_DEFINE0\s*\(\s*(a-zA-Z0-9_)*\s*\)/) { - if ($prototype =~ m/SYSCALL_DEFINE0/) { - $void = 1; -## $prototype = "long sys_$1(void)"; - } - - $prototype =~ s/SYSCALL_DEFINE.*\(/long sys_/; # fix return type & func name - if ($prototype =~ m/long (sys_.*?),/) { - $prototype =~ s/,/\(/; - } elsif ($void) { - $prototype =~ s/\)/\(void\)/; - } - - # now delete all of the odd-number commas in $prototype - # so that arg types & arg names don't have a comma between them - my $count = 0; - my $len = length($prototype); - if ($void) { - $len = 0; # skip the for-loop - } - for (my $ix = 0; $ix < $len; $ix++) { - if (substr($prototype, $ix, 1) eq ',') { - $count++; - if ($count % 2 == 1) { - substr($prototype, $ix, 1) = ' '; - } - } - } -} - -sub process_proto_function($$) { - my $x = shift; - my $file = shift; - - $x =~ s@\/\/.*$@@gos; # strip C99-style comments to end of line - - if ($x =~ m#\s*/\*\s+MACDOC\s*#io || ($x =~ /^#/ && $x !~ /^#\s*define/)) { - # do nothing - } - elsif ($x =~ /([^\{]*)/) { - $prototype .= $1; - } - - if (($x =~ /\{/) || ($x =~ /\#\s*define/) || ($x =~ /;/)) { - $prototype =~ s@/\*.*?\*/@@gos; # strip comments. - $prototype =~ s@[\r\n]+@ @gos; # strip newlines/cr's. - $prototype =~ s@^\s+@@gos; # strip leading spaces - - # Handle prototypes for function pointers like: - # int (*pcs_config)(struct foo) - $prototype =~ s@^(\S+\s+)\(\s*\*(\S+)\)@$1$2@gos; - - if ($prototype =~ /SYSCALL_DEFINE/) { - syscall_munge(); - } - if ($prototype =~ /TRACE_EVENT/ || $prototype =~ /DEFINE_EVENT/ || - $prototype =~ /DEFINE_SINGLE_EVENT/) - { - tracepoint_munge($file); - } - dump_function($prototype, $file); - reset_state(); - } -} - -sub process_proto_type($$) { - my $x = shift; - my $file = shift; - - $x =~ s@[\r\n]+@ @gos; # strip newlines/cr's. - $x =~ s@^\s+@@gos; # strip leading spaces - $x =~ s@\s+$@@gos; # strip trailing spaces - $x =~ s@\/\/.*$@@gos; # strip C99-style comments to end of line - - if ($x =~ /^#/) { - # To distinguish preprocessor directive from regular declaration later. - $x .= ";"; - } - - while (1) { - if ( $x =~ /([^\{\};]*)([\{\};])(.*)/ ) { - if( length $prototype ) { - $prototype .= " " - } - $prototype .= $1 . $2; - ($2 eq '{') && $brcount++; - ($2 eq '}') && $brcount--; - if (($2 eq ';') && ($brcount == 0)) { - dump_declaration($prototype, $file); - reset_state(); - last; - } - $x = $3; - } else { - $prototype .= $x; - last; - } - } -} - - -sub map_filename($) { - my $file; - my ($orig_file) = @_; - - if (defined($ENV{'SRCTREE'})) { - $file = "$ENV{'SRCTREE'}" . "/" . $orig_file; - } else { - $file = $orig_file; - } - - if (defined($source_map{$file})) { - $file = $source_map{$file}; - } - - return $file; -} - -sub process_export_file($) { - my ($orig_file) = @_; - my $file = map_filename($orig_file); - - if (!open(IN,"<$file")) { - print STDERR "Error: Cannot open file $file\n"; - ++$errors; - return; - } - - while () { - if (/$export_symbol/) { - next if (defined($nosymbol_table{$2})); - $function_table{$2} = 1; - } - } - - close(IN); -} - -# -# Parsers for the various processing states. -# -# STATE_NORMAL: looking for the /** to begin everything. -# -sub process_normal() { - if (/$doc_start/o) { - $state = STATE_NAME; # next line is always the function name - $in_doc_sect = 0; - $declaration_start_line = $. + 1; - } -} - -# -# STATE_NAME: Looking for the "name - description" line -# -sub process_name($$) { - my $file = shift; - my $identifier; - my $descr; - - if (/$doc_block/o) { - $state = STATE_DOCBLOCK; - $contents = ""; - $new_start_line = $.; - - if ( $1 eq "" ) { - $section = $section_intro; - } else { - $section = $1; - } - } - elsif (/$doc_decl/o) { - $identifier = $1; - if (/\s*([\w\s]+?)(\s*-|:)/) { - $identifier = $1; - } - - $state = STATE_BODY; - # if there's no @param blocks need to set up default section - # here - $contents = ""; - $section = $section_default; - $new_start_line = $. + 1; - if (/[-:](.*)/) { - # strip leading/trailing/multiple spaces - $descr= $1; - $descr =~ s/^\s*//; - $descr =~ s/\s*$//; - $descr =~ s/\s+/ /g; - $declaration_purpose = $descr; - $state = STATE_BODY_MAYBE; - } else { - $declaration_purpose = ""; - } - - if (($declaration_purpose eq "") && $verbose) { - print STDERR "${file}:$.: warning: missing initial short description on line:\n"; - print STDERR $_; - ++$warnings; - } - - if ($identifier =~ m/^struct\b/) { - $decl_type = 'struct'; - } elsif ($identifier =~ m/^union\b/) { - $decl_type = 'union'; - } elsif ($identifier =~ m/^enum\b/) { - $decl_type = 'enum'; - } elsif ($identifier =~ m/^typedef\b/) { - $decl_type = 'typedef'; - } else { - $decl_type = 'function'; - } - - if ($verbose) { - print STDERR "${file}:$.: info: Scanning doc for $identifier\n"; - } - } else { - print STDERR "${file}:$.: warning: Cannot understand $_ on line $.", - " - I thought it was a doc line\n"; - ++$warnings; - $state = STATE_NORMAL; - } -} - - -# -# STATE_BODY and STATE_BODY_MAYBE: the bulk of a kerneldoc comment. -# -sub process_body($$) { - my $file = shift; - - # Until all named variable macro parameters are - # documented using the bare name (`x`) rather than with - # dots (`x...`), strip the dots: - if ($section =~ /\w\.\.\.$/) { - $section =~ s/\.\.\.$//; - - if ($verbose) { - print STDERR "${file}:$.: warning: Variable macro arguments should be documented without dots\n"; - ++$warnings; - } - } - - if ($state == STATE_BODY_WITH_BLANK_LINE && /^\s*\*\s?\S/) { - dump_section($file, $section, $contents); - $section = $section_default; - $new_start_line = $.; - $contents = ""; - } - - if (/$doc_sect/i) { # case insensitive for supported section names - $newsection = $1; - $newcontents = $2; - - # map the supported section names to the canonical names - if ($newsection =~ m/^description$/i) { - $newsection = $section_default; - } elsif ($newsection =~ m/^context$/i) { - $newsection = $section_context; - } elsif ($newsection =~ m/^returns?$/i) { - $newsection = $section_return; - } elsif ($newsection =~ m/^\@return$/) { - # special: @return is a section, not a param description - $newsection = $section_return; - } - - if (($contents ne "") && ($contents ne "\n")) { - if (!$in_doc_sect && $verbose) { - print STDERR "${file}:$.: warning: contents before sections\n"; - ++$warnings; - } - dump_section($file, $section, $contents); - $section = $section_default; - } - - $in_doc_sect = 1; - $state = STATE_BODY; - $contents = $newcontents; - $new_start_line = $.; - while (substr($contents, 0, 1) eq " ") { - $contents = substr($contents, 1); - } - if ($contents ne "") { - $contents .= "\n"; - } - $section = $newsection; - $leading_space = undef; - } elsif (/$doc_end/) { - if (($contents ne "") && ($contents ne "\n")) { - dump_section($file, $section, $contents); - $section = $section_default; - $contents = ""; - } - # look for doc_com + + doc_end: - if ($_ =~ m'\s*\*\s*[a-zA-Z_0-9:\.]+\*/') { - print STDERR "${file}:$.: warning: suspicious ending line: $_"; - ++$warnings; - } - - $prototype = ""; - $state = STATE_PROTO; - $brcount = 0; - $new_start_line = $. + 1; - } elsif (/$doc_content/) { - if ($1 eq "") { - if ($section eq $section_context) { - dump_section($file, $section, $contents); - $section = $section_default; - $contents = ""; - $new_start_line = $.; - $state = STATE_BODY; - } else { - if ($section ne $section_default) { - $state = STATE_BODY_WITH_BLANK_LINE; - } else { - $state = STATE_BODY; - } - $contents .= "\n"; - } - } elsif ($state == STATE_BODY_MAYBE) { - # Continued declaration purpose - chomp($declaration_purpose); - $declaration_purpose .= " " . $1; - $declaration_purpose =~ s/\s+/ /g; - } else { - my $cont = $1; - if ($section =~ m/^@/ || $section eq $section_context) { - if (!defined $leading_space) { - if ($cont =~ m/^(\s+)/) { - $leading_space = $1; - } else { - $leading_space = ""; - } - } - $cont =~ s/^$leading_space//; - } - $contents .= $cont . "\n"; - } - } else { - # i dont know - bad line? ignore. - print STDERR "${file}:$.: warning: bad line: $_"; - ++$warnings; - } -} - - -# -# STATE_PROTO: reading a function/whatever prototype. -# -sub process_proto($$) { - my $file = shift; - - if (/$doc_inline_oneline/) { - $section = $1; - $contents = $2; - if ($contents ne "") { - $contents .= "\n"; - dump_section($file, $section, $contents); - $section = $section_default; - $contents = ""; - } - } elsif (/$doc_inline_start/) { - $state = STATE_INLINE; - $inline_doc_state = STATE_INLINE_NAME; - } elsif ($decl_type eq 'function') { - process_proto_function($_, $file); - } else { - process_proto_type($_, $file); - } -} - -# -# STATE_DOCBLOCK: within a DOC: block. -# -sub process_docblock($$) { - my $file = shift; - - if (/$doc_end/) { - dump_doc_section($file, $section, $contents); - $section = $section_default; - $contents = ""; - $function = ""; - %parameterdescs = (); - %parametertypes = (); - @parameterlist = (); - %sections = (); - @sectionlist = (); - $prototype = ""; - $state = STATE_NORMAL; - } elsif (/$doc_content/) { - if ( $1 eq "" ) { - $contents .= $blankline; - } else { - $contents .= $1 . "\n"; - } - } -} - -# -# STATE_INLINE: docbook comments within a prototype. -# -sub process_inline($$) { - my $file = shift; - - # First line (state 1) needs to be a @parameter - if ($inline_doc_state == STATE_INLINE_NAME && /$doc_inline_sect/o) { - $section = $1; - $contents = $2; - $new_start_line = $.; - if ($contents ne "") { - while (substr($contents, 0, 1) eq " ") { - $contents = substr($contents, 1); - } - $contents .= "\n"; - } - $inline_doc_state = STATE_INLINE_TEXT; - # Documentation block end */ - } elsif (/$doc_inline_end/) { - if (($contents ne "") && ($contents ne "\n")) { - dump_section($file, $section, $contents); - $section = $section_default; - $contents = ""; - } - $state = STATE_PROTO; - $inline_doc_state = STATE_INLINE_NA; - # Regular text - } elsif (/$doc_content/) { - if ($inline_doc_state == STATE_INLINE_TEXT) { - $contents .= $1 . "\n"; - # nuke leading blank lines - if ($contents =~ /^\s*$/) { - $contents = ""; - } - } elsif ($inline_doc_state == STATE_INLINE_NAME) { - $inline_doc_state = STATE_INLINE_ERROR; - print STDERR "${file}:$.: warning: "; - print STDERR "Incorrect use of kernel-doc format: $_"; - ++$warnings; - } - } -} - - -sub process_file($) { - my $file; - my $initial_section_counter = $section_counter; - my ($orig_file) = @_; - - $file = map_filename($orig_file); - - if (!open(IN_FILE,"<$file")) { - print STDERR "Error: Cannot open file $file\n"; - ++$errors; - return; - } - - $. = 1; - - $section_counter = 0; - while () { - while (s/\\\s*$//) { - $_ .= ; - } - # Replace tabs by spaces - while ($_ =~ s/\t+/' ' x (length($&) * 8 - length($`) % 8)/e) {}; - # Hand this line to the appropriate state handler - if ($state == STATE_NORMAL) { - process_normal(); - } elsif ($state == STATE_NAME) { - process_name($file, $_); - } elsif ($state == STATE_BODY || $state == STATE_BODY_MAYBE || - $state == STATE_BODY_WITH_BLANK_LINE) { - process_body($file, $_); - } elsif ($state == STATE_INLINE) { # scanning for inline parameters - process_inline($file, $_); - } elsif ($state == STATE_PROTO) { - process_proto($file, $_); - } elsif ($state == STATE_DOCBLOCK) { - process_docblock($file, $_); - } - } - - # Make sure we got something interesting. - if ($initial_section_counter == $section_counter && $ - output_mode ne "none") { - if ($output_selection == OUTPUT_INCLUDE) { - print STDERR "${file}:1: warning: '$_' not found\n" - for keys %function_table; - } - else { - print STDERR "${file}:1: warning: no structured comments found\n"; - } - } - close IN_FILE; -} - - -if ($output_mode eq "rst") { - get_sphinx_version() if (!$sphinx_major); -} - -$kernelversion = get_kernel_version(); - -# generate a sequence of code that will splice in highlighting information -# using the s// operator. -for (my $k = 0; $k < @highlights; $k++) { - my $pattern = $highlights[$k][0]; - my $result = $highlights[$k][1]; -# print STDERR "scanning pattern:$pattern, highlight:($result)\n"; - $dohighlight .= "\$contents =~ s:$pattern:$result:gs;\n"; -} - -# Read the file that maps relative names to absolute names for -# separate source and object directories and for shadow trees. -if (open(SOURCE_MAP, "<.tmp_filelist.txt")) { - my ($relname, $absname); - while() { - chop(); - ($relname, $absname) = (split())[0..1]; - $relname =~ s:^/+::; - $source_map{$relname} = $absname; - } - close(SOURCE_MAP); -} - -if ($output_selection == OUTPUT_EXPORTED || - $output_selection == OUTPUT_INTERNAL) { - - push(@export_file_list, @ARGV); - - foreach (@export_file_list) { - chomp; - process_export_file($_); - } -} - -foreach (@ARGV) { - chomp; - process_file($_); -} -if ($verbose && $errors) { - print STDERR "$errors errors\n"; -} -if ($verbose && $warnings) { - print STDERR "$warnings warnings\n"; -} - -if ($Werror && $warnings) { - print STDERR "$warnings warnings as Errors\n"; - exit($warnings); -} else { - exit($output_mode eq "none" ? 0 : $errors) -} From e0ca100425853aa362acefeb027800d952fb222d Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 14 Aug 2025 18:13:23 +0100 Subject: [PATCH 23/32] MAINTAINERS: Put kernel-doc under the "docs build machinery" section We never had a MAINTAINERS entry for the old kernel-doc script; add the files for the new Python kernel-doc under "Sphinx documentation configuration and build machinery", as the most appropriate subsection. Mauro has kindly volunteered to help with maintenance/review of this area of the codebase, so add him as a maintainer. Signed-off-by: Peter Maydell Reviewed-by: Paolo Bonzini Reviewed-by: Mauro Carvalho Chehab Message-id: 20250814171324.1614516-9-peter.maydell@linaro.org --- MAINTAINERS | 3 +++ 1 file changed, 3 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 8f074e4371..8147fff352 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -4441,6 +4441,7 @@ F: po/*.po Sphinx documentation configuration and build machinery M: John Snow M: Peter Maydell +M: Mauro Carvalho Chehab S: Maintained F: docs/conf.py F: docs/*/conf.py @@ -4449,6 +4450,8 @@ F: docs/sphinx/ F: docs/_templates/ F: docs/devel/docs.rst F: docs/devel/qapi-domain.rst +F: scripts/kernel-doc +F: scripts/lib/kdoc/ Rust build system integration M: Manos Pitsidianakis From c2fae597099ef6ff81ca63d69ee28eddb982d894 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Tue, 19 Aug 2025 15:56:58 +0100 Subject: [PATCH 24/32] target/arm: Correct condition of aa64_atomics feature function The ARMv8.1-Atomics feature (renamed FEAT_LSE in more modern versions of the Arm ARM) has always ben indicated by ID_AA64ISAR0.ATOMIC being 0b0010 or greater; 0b0001 is a reserved unused value. We were incorrectly checking for != 0; this had no harmful effects because all the CPUs set their value for this field to either 0 (for not having the feature) or 2 (if they do have it), but it's better to match what the architecture specifies here. Signed-off-by: Peter Maydell Reviewed-by: Richard Henderson Message-id: 20250819145659.2165160-1-peter.maydell@linaro.org --- target/arm/cpu-features.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index 41511d0835..d48754bcf2 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -408,7 +408,7 @@ static inline bool isar_feature_aa64_crc32(const ARMISARegisters *id) static inline bool isar_feature_aa64_atomics(const ARMISARegisters *id) { - return FIELD_EX64_IDREG(id, ID_AA64ISAR0, ATOMIC) != 0; + return FIELD_EX64_IDREG(id, ID_AA64ISAR0, ATOMIC) >= 2; } static inline bool isar_feature_aa64_rdm(const ARMISARegisters *id) From 1748c0d59228c7790940d8be381df1c3108022b1 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:47 +1000 Subject: [PATCH 25/32] qemu/atomic: Finish renaming atomic128-cas.h headers The aarch64 header was not renamed with the others, meaning it was skipped in favor of the generic version. Cc: qemu-stable@nongnu.org Fixes: 15606965400b ("qemu/atomic: Rename atomic128-cas.h headers using .h.inc suffix") Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- .../include/aarch64/host/{atomic128-cas.h => atomic128-cas.h.inc} | 0 1 file changed, 0 insertions(+), 0 deletions(-) rename host/include/aarch64/host/{atomic128-cas.h => atomic128-cas.h.inc} (100%) diff --git a/host/include/aarch64/host/atomic128-cas.h b/host/include/aarch64/host/atomic128-cas.h.inc similarity index 100% rename from host/include/aarch64/host/atomic128-cas.h rename to host/include/aarch64/host/atomic128-cas.h.inc From 1d5e88e7b5b83130deb312d05de11267baac4c05 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:48 +1000 Subject: [PATCH 26/32] qemu/atomic: Add atomic16 primitives for xchg, fetch_and, fetch_or Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- host/include/aarch64/host/atomic128-cas.h.inc | 57 +++++++++++ host/include/generic/host/atomic128-cas.h.inc | 96 +++++++++++++++++++ 2 files changed, 153 insertions(+) diff --git a/host/include/aarch64/host/atomic128-cas.h.inc b/host/include/aarch64/host/atomic128-cas.h.inc index 991da4ef54..aec27df182 100644 --- a/host/include/aarch64/host/atomic128-cas.h.inc +++ b/host/include/aarch64/host/atomic128-cas.h.inc @@ -38,6 +38,63 @@ static inline Int128 atomic16_cmpxchg(Int128 *ptr, Int128 cmp, Int128 new) return int128_make128(oldl, oldh); } +static inline Int128 atomic16_xchg(Int128 *ptr, Int128 new) +{ + uint64_t newl = int128_getlo(new), newh = int128_gethi(new); + uint64_t oldl, oldh; + uint32_t tmp; + + asm("0: ldaxp %[oldl], %[oldh], %[mem]\n\t" + "stlxp %w[tmp], %[newl], %[newh], %[mem]\n\t" + "cbnz %w[tmp], 0b" + : [mem] "+m"(*ptr), [tmp] "=&r"(tmp), + [oldl] "=&r"(oldl), [oldh] "=&r"(oldh) + : [newl] "r"(newl), [newh] "r"(newh) + : "memory"); + + return int128_make128(oldl, oldh); +} + +static inline Int128 atomic16_fetch_and(Int128 *ptr, Int128 new) +{ + uint64_t newl = int128_getlo(new), newh = int128_gethi(new); + uint64_t oldl, oldh, tmpl, tmph; + uint32_t tmp; + + asm("0: ldaxp %[oldl], %[oldh], %[mem]\n\t" + "and %[tmpl], %[oldl], %[newl]\n\t" + "and %[tmph], %[oldh], %[newh]\n\t" + "stlxp %w[tmp], %[tmpl], %[tmph], %[mem]\n\t" + "cbnz %w[tmp], 0b" + : [mem] "+m"(*ptr), [tmp] "=&r"(tmp), + [oldl] "=&r"(oldl), [oldh] "=&r"(oldh) + : [newl] "r"(newl), [newh] "r"(newh), + [tmpl] "r"(tmpl), [tmph] "r"(tmph) + : "memory"); + + return int128_make128(oldl, oldh); +} + +static inline Int128 atomic16_fetch_or(Int128 *ptr, Int128 new) +{ + uint64_t newl = int128_getlo(new), newh = int128_gethi(new); + uint64_t oldl, oldh, tmpl, tmph; + uint32_t tmp; + + asm("0: ldaxp %[oldl], %[oldh], %[mem]\n\t" + "orr %[tmpl], %[oldl], %[newl]\n\t" + "orr %[tmph], %[oldh], %[newh]\n\t" + "stlxp %w[tmp], %[tmpl], %[tmph], %[mem]\n\t" + "cbnz %w[tmp], 0b" + : [mem] "+m"(*ptr), [tmp] "=&r"(tmp), + [oldl] "=&r"(oldl), [oldh] "=&r"(oldh) + : [newl] "r"(newl), [newh] "r"(newh), + [tmpl] "r"(tmpl), [tmph] "r"(tmph) + : "memory"); + + return int128_make128(oldl, oldh); +} + # define CONFIG_CMPXCHG128 1 # define HAVE_CMPXCHG128 1 #endif diff --git a/host/include/generic/host/atomic128-cas.h.inc b/host/include/generic/host/atomic128-cas.h.inc index 6b40cc2271..990162c56f 100644 --- a/host/include/generic/host/atomic128-cas.h.inc +++ b/host/include/generic/host/atomic128-cas.h.inc @@ -23,6 +23,51 @@ atomic16_cmpxchg(Int128 *ptr, Int128 cmp, Int128 new) r.i = qatomic_cmpxchg__nocheck(ptr_align, c.i, n.i); return r.s; } + +/* + * Since we're looping anyway, use weak compare and swap. + * If the host supports weak, this will eliminate a second loop hidden + * within the atomic operation itself; otherwise the weak parameter is + * ignored. + */ +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_xchg(Int128 *ptr, Int128 new) +{ + __int128_t *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128 old = *ptr_align; + + while (!__atomic_compare_exchange_n(ptr_align, &old, new, true, + __ATOMIC_SEQ_CST, 0)) { + continue; + } + return old; +} + +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_fetch_and(Int128 *ptr, Int128 val) +{ + __int128_t *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128 old = *ptr_align; + + while (!__atomic_compare_exchange_n(ptr_align, &old, old & val, true, + __ATOMIC_SEQ_CST, 0)) { + continue; + } + return old; +} + +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_fetch_or(Int128 *ptr, Int128 val) +{ + __int128_t *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128 old = *ptr_align; + + while (!__atomic_compare_exchange_n(ptr_align, &old, old | val, true, + __ATOMIC_SEQ_CST, 0)) { + continue; + } + return old; +} # define HAVE_CMPXCHG128 1 #elif defined(CONFIG_CMPXCHG128) static inline Int128 ATTRIBUTE_ATOMIC128_OPT @@ -36,6 +81,57 @@ atomic16_cmpxchg(Int128 *ptr, Int128 cmp, Int128 new) r.i = __sync_val_compare_and_swap_16(ptr_align, c.i, n.i); return r.s; } + +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_xchg(Int128 *ptr, Int128 new) +{ + Int128Aligned *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128Alias o, n; + + n.s = new; + o.s = *ptr_align; + while (1) { + __int128 c = __sync_val_compare_and_swap_16(ptr_align, o.i, n.i); + if (c == o.i) { + return o.s; + } + o.i = c; + } +} + +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_fetch_and(Int128 *ptr, Int128 val) +{ + Int128Aligned *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128Alias o, v; + + v.s = val; + o.s = *ptr_align; + while (1) { + __int128 c = __sync_val_compare_and_swap_16(ptr_align, o.i, o.i & v.i); + if (c == o.i) { + return o.s; + } + o.i = c; + } +} + +static inline Int128 ATTRIBUTE_ATOMIC128_OPT +atomic16_fetch_or(Int128 *ptr, Int128 val) +{ + Int128Aligned *ptr_align = __builtin_assume_aligned(ptr, 16); + Int128Alias o, v; + + v.s = val; + o.s = *ptr_align; + while (1) { + __int128 c = __sync_val_compare_and_swap_16(ptr_align, o.i, o.i | v.i); + if (c == o.i) { + return o.s; + } + o.i = c; + } +} # define HAVE_CMPXCHG128 1 #else /* Fallback definition that must be optimized away, or error. */ From 33aefd187eaff71dbc686c43f7acfdd0a81c7de4 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:49 +1000 Subject: [PATCH 27/32] accel/tcg: Add cpu_atomic_*_mmu for 16-byte xchg, fetch_and, fetch_or Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- accel/tcg/atomic_template.h | 80 +++++++++++++++++++++++++++-- include/accel/tcg/cpu-ldst-common.h | 13 +++-- 2 files changed, 86 insertions(+), 7 deletions(-) diff --git a/accel/tcg/atomic_template.h b/accel/tcg/atomic_template.h index 08a475c10c..ae5203b439 100644 --- a/accel/tcg/atomic_template.h +++ b/accel/tcg/atomic_template.h @@ -100,7 +100,6 @@ ABI_TYPE ATOMIC_NAME(cmpxchg)(CPUArchState *env, vaddr addr, return ret; } -#if DATA_SIZE < 16 ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, MemOpIdx oi, uintptr_t retaddr) { @@ -108,7 +107,11 @@ ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, DATA_SIZE, retaddr); DATA_TYPE ret; +#if DATA_SIZE == 16 + ret = atomic16_xchg(haddr, val); +#else ret = qatomic_xchg__nocheck(haddr, val); +#endif ATOMIC_MMU_CLEANUP; atomic_trace_rmw_post(env, addr, VALUE_LOW(ret), @@ -119,6 +122,39 @@ ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, return ret; } +#if DATA_SIZE == 16 +ABI_TYPE ATOMIC_NAME(fetch_and)(CPUArchState *env, vaddr addr, ABI_TYPE val, + MemOpIdx oi, uintptr_t retaddr) +{ + DATA_TYPE *haddr = atomic_mmu_lookup(env_cpu(env), addr, oi, + DATA_SIZE, retaddr); + DATA_TYPE ret = atomic16_fetch_and(haddr, val); + ATOMIC_MMU_CLEANUP; + atomic_trace_rmw_post(env, addr, + VALUE_LOW(ret), + VALUE_HIGH(ret), + VALUE_LOW(val), + VALUE_HIGH(val), + oi); + return ret; +} + +ABI_TYPE ATOMIC_NAME(fetch_or)(CPUArchState *env, vaddr addr, ABI_TYPE val, + MemOpIdx oi, uintptr_t retaddr) +{ + DATA_TYPE *haddr = atomic_mmu_lookup(env_cpu(env), addr, oi, + DATA_SIZE, retaddr); + DATA_TYPE ret = atomic16_fetch_or(haddr, val); + ATOMIC_MMU_CLEANUP; + atomic_trace_rmw_post(env, addr, + VALUE_LOW(ret), + VALUE_HIGH(ret), + VALUE_LOW(val), + VALUE_HIGH(val), + oi); + return ret; +} +#else #define GEN_ATOMIC_HELPER(X) \ ABI_TYPE ATOMIC_NAME(X)(CPUArchState *env, vaddr addr, \ ABI_TYPE val, MemOpIdx oi, uintptr_t retaddr) \ @@ -188,7 +224,7 @@ GEN_ATOMIC_HELPER_FN(smax_fetch, MAX, SDATA_TYPE, new) GEN_ATOMIC_HELPER_FN(umax_fetch, MAX, DATA_TYPE, new) #undef GEN_ATOMIC_HELPER_FN -#endif /* DATA SIZE < 16 */ +#endif /* DATA SIZE == 16 */ #undef END @@ -225,7 +261,6 @@ ABI_TYPE ATOMIC_NAME(cmpxchg)(CPUArchState *env, vaddr addr, return BSWAP(ret); } -#if DATA_SIZE < 16 ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, MemOpIdx oi, uintptr_t retaddr) { @@ -233,7 +268,11 @@ ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, DATA_SIZE, retaddr); ABI_TYPE ret; +#if DATA_SIZE == 16 + ret = atomic16_xchg(haddr, BSWAP(val)); +#else ret = qatomic_xchg__nocheck(haddr, BSWAP(val)); +#endif ATOMIC_MMU_CLEANUP; atomic_trace_rmw_post(env, addr, VALUE_LOW(ret), @@ -244,6 +283,39 @@ ABI_TYPE ATOMIC_NAME(xchg)(CPUArchState *env, vaddr addr, ABI_TYPE val, return BSWAP(ret); } +#if DATA_SIZE == 16 +ABI_TYPE ATOMIC_NAME(fetch_and)(CPUArchState *env, vaddr addr, ABI_TYPE val, + MemOpIdx oi, uintptr_t retaddr) +{ + DATA_TYPE *haddr = atomic_mmu_lookup(env_cpu(env), addr, oi, + DATA_SIZE, retaddr); + DATA_TYPE ret = atomic16_fetch_and(haddr, BSWAP(val)); + ATOMIC_MMU_CLEANUP; + atomic_trace_rmw_post(env, addr, + VALUE_LOW(ret), + VALUE_HIGH(ret), + VALUE_LOW(val), + VALUE_HIGH(val), + oi); + return BSWAP(ret); +} + +ABI_TYPE ATOMIC_NAME(fetch_or)(CPUArchState *env, vaddr addr, ABI_TYPE val, + MemOpIdx oi, uintptr_t retaddr) +{ + DATA_TYPE *haddr = atomic_mmu_lookup(env_cpu(env), addr, oi, + DATA_SIZE, retaddr); + DATA_TYPE ret = atomic16_fetch_or(haddr, BSWAP(val)); + ATOMIC_MMU_CLEANUP; + atomic_trace_rmw_post(env, addr, + VALUE_LOW(ret), + VALUE_HIGH(ret), + VALUE_LOW(val), + VALUE_HIGH(val), + oi); + return BSWAP(ret); +} +#else #define GEN_ATOMIC_HELPER(X) \ ABI_TYPE ATOMIC_NAME(X)(CPUArchState *env, vaddr addr, \ ABI_TYPE val, MemOpIdx oi, uintptr_t retaddr) \ @@ -317,7 +389,7 @@ GEN_ATOMIC_HELPER_FN(add_fetch, ADD, DATA_TYPE, new) #undef ADD #undef GEN_ATOMIC_HELPER_FN -#endif /* DATA_SIZE < 16 */ +#endif /* DATA_SIZE == 16 */ #undef END #endif /* DATA_SIZE > 1 */ diff --git a/include/accel/tcg/cpu-ldst-common.h b/include/accel/tcg/cpu-ldst-common.h index 8bf17c2fab..17a3250ded 100644 --- a/include/accel/tcg/cpu-ldst-common.h +++ b/include/accel/tcg/cpu-ldst-common.h @@ -100,9 +100,6 @@ GEN_ATOMIC_HELPER_ALL(umax_fetch) GEN_ATOMIC_HELPER_ALL(xchg) -#undef GEN_ATOMIC_HELPER_ALL -#undef GEN_ATOMIC_HELPER - Int128 cpu_atomic_cmpxchgo_le_mmu(CPUArchState *env, vaddr addr, Int128 cmpv, Int128 newv, MemOpIdx oi, uintptr_t retaddr); @@ -110,6 +107,16 @@ Int128 cpu_atomic_cmpxchgo_be_mmu(CPUArchState *env, vaddr addr, Int128 cmpv, Int128 newv, MemOpIdx oi, uintptr_t retaddr); +GEN_ATOMIC_HELPER(xchg, Int128, o_le) +GEN_ATOMIC_HELPER(xchg, Int128, o_be) +GEN_ATOMIC_HELPER(fetch_and, Int128, o_le) +GEN_ATOMIC_HELPER(fetch_and, Int128, o_be) +GEN_ATOMIC_HELPER(fetch_or, Int128, o_le) +GEN_ATOMIC_HELPER(fetch_or, Int128, o_be) + +#undef GEN_ATOMIC_HELPER_ALL +#undef GEN_ATOMIC_HELPER + uint8_t cpu_ldb_code_mmu(CPUArchState *env, vaddr addr, MemOpIdx oi, uintptr_t ra); uint16_t cpu_ldw_code_mmu(CPUArchState *env, vaddr addr, From 092ac2481a4301d0282227bb4ee8641b3f39e437 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:50 +1000 Subject: [PATCH 28/32] tcg: Add tcg_gen_atomic_{xchg,fetch_and,fetch_or}_i128 Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-5-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- accel/tcg/atomic_common.c.inc | 9 ++++ accel/tcg/tcg-runtime.h | 12 +++++ include/tcg/tcg-op-common.h | 7 +++ include/tcg/tcg-op.h | 3 ++ tcg/tcg-op-ldst.c | 97 +++++++++++++++++++++++++++++++++-- 5 files changed, 125 insertions(+), 3 deletions(-) diff --git a/accel/tcg/atomic_common.c.inc b/accel/tcg/atomic_common.c.inc index 6056598c23..bca93a0ac4 100644 --- a/accel/tcg/atomic_common.c.inc +++ b/accel/tcg/atomic_common.c.inc @@ -122,5 +122,14 @@ GEN_ATOMIC_HELPERS(umax_fetch) GEN_ATOMIC_HELPERS(xchg) +#if HAVE_CMPXCHG128 +ATOMIC_HELPER(xchgo_be, Int128) +ATOMIC_HELPER(xchgo_le, Int128) +ATOMIC_HELPER(fetch_ando_be, Int128) +ATOMIC_HELPER(fetch_ando_le, Int128) +ATOMIC_HELPER(fetch_oro_be, Int128) +ATOMIC_HELPER(fetch_oro_le, Int128) +#endif + #undef ATOMIC_HELPER #undef GEN_ATOMIC_HELPERS diff --git a/accel/tcg/tcg-runtime.h b/accel/tcg/tcg-runtime.h index c23b5e66c4..8436599b9f 100644 --- a/accel/tcg/tcg-runtime.h +++ b/accel/tcg/tcg-runtime.h @@ -63,6 +63,18 @@ DEF_HELPER_FLAGS_5(atomic_cmpxchgo_be, TCG_CALL_NO_WG, i128, env, i64, i128, i128, i32) DEF_HELPER_FLAGS_5(atomic_cmpxchgo_le, TCG_CALL_NO_WG, i128, env, i64, i128, i128, i32) +DEF_HELPER_FLAGS_4(atomic_xchgo_be, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) +DEF_HELPER_FLAGS_4(atomic_xchgo_le, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) +DEF_HELPER_FLAGS_4(atomic_fetch_ando_be, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) +DEF_HELPER_FLAGS_4(atomic_fetch_ando_le, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) +DEF_HELPER_FLAGS_4(atomic_fetch_oro_be, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) +DEF_HELPER_FLAGS_4(atomic_fetch_oro_le, TCG_CALL_NO_WG, + i128, env, i64, i128, i32) #endif DEF_HELPER_FLAGS_5(nonatomic_cmpxchgo, TCG_CALL_NO_WG, diff --git a/include/tcg/tcg-op-common.h b/include/tcg/tcg-op-common.h index e1071adebf..f752ef440b 100644 --- a/include/tcg/tcg-op-common.h +++ b/include/tcg/tcg-op-common.h @@ -344,6 +344,8 @@ void tcg_gen_atomic_xchg_i32_chk(TCGv_i32, TCGTemp *, TCGv_i32, TCGArg, MemOp, TCGType); void tcg_gen_atomic_xchg_i64_chk(TCGv_i64, TCGTemp *, TCGv_i64, TCGArg, MemOp, TCGType); +void tcg_gen_atomic_xchg_i128_chk(TCGv_i128, TCGTemp *, TCGv_i128, + TCGArg, MemOp, TCGType); void tcg_gen_atomic_fetch_add_i32_chk(TCGv_i32, TCGTemp *, TCGv_i32, TCGArg, MemOp, TCGType); @@ -411,6 +413,11 @@ void tcg_gen_atomic_umax_fetch_i32_chk(TCGv_i32, TCGTemp *, TCGv_i32, void tcg_gen_atomic_umax_fetch_i64_chk(TCGv_i64, TCGTemp *, TCGv_i64, TCGArg, MemOp, TCGType); +void tcg_gen_atomic_fetch_and_i128_chk(TCGv_i128, TCGTemp *, TCGv_i128, + TCGArg, MemOp, TCGType); +void tcg_gen_atomic_fetch_or_i128_chk(TCGv_i128, TCGTemp *, TCGv_i128, + TCGArg, MemOp, TCGType); + /* Vector ops */ void tcg_gen_mov_vec(TCGv_vec, TCGv_vec); diff --git a/include/tcg/tcg-op.h b/include/tcg/tcg-op.h index c912578fdd..232733cb71 100644 --- a/include/tcg/tcg-op.h +++ b/include/tcg/tcg-op.h @@ -134,13 +134,16 @@ DEF_ATOMIC3(tcg_gen_nonatomic_cmpxchg, i128) DEF_ATOMIC2(tcg_gen_atomic_xchg, i32) DEF_ATOMIC2(tcg_gen_atomic_xchg, i64) +DEF_ATOMIC2(tcg_gen_atomic_xchg, i128) DEF_ATOMIC2(tcg_gen_atomic_fetch_add, i32) DEF_ATOMIC2(tcg_gen_atomic_fetch_add, i64) DEF_ATOMIC2(tcg_gen_atomic_fetch_and, i32) DEF_ATOMIC2(tcg_gen_atomic_fetch_and, i64) +DEF_ATOMIC2(tcg_gen_atomic_fetch_and, i128) DEF_ATOMIC2(tcg_gen_atomic_fetch_or, i32) DEF_ATOMIC2(tcg_gen_atomic_fetch_or, i64) +DEF_ATOMIC2(tcg_gen_atomic_fetch_or, i128) DEF_ATOMIC2(tcg_gen_atomic_fetch_xor, i32) DEF_ATOMIC2(tcg_gen_atomic_fetch_xor, i64) DEF_ATOMIC2(tcg_gen_atomic_fetch_smin, i32) diff --git a/tcg/tcg-op-ldst.c b/tcg/tcg-op-ldst.c index 548496002d..67c15fd4d0 100644 --- a/tcg/tcg-op-ldst.c +++ b/tcg/tcg-op-ldst.c @@ -801,6 +801,8 @@ typedef void (*gen_atomic_op_i32)(TCGv_i32, TCGv_env, TCGv_i64, TCGv_i32, TCGv_i32); typedef void (*gen_atomic_op_i64)(TCGv_i64, TCGv_env, TCGv_i64, TCGv_i64, TCGv_i32); +typedef void (*gen_atomic_op_i128)(TCGv_i128, TCGv_env, TCGv_i64, + TCGv_i128, TCGv_i32); #ifdef CONFIG_ATOMIC64 # define WITH_ATOMIC64(X) X, @@ -1201,6 +1203,94 @@ static void do_atomic_op_i64(TCGv_i64 ret, TCGTemp *addr, TCGv_i64 val, } } +static void do_nonatomic_op_i128(TCGv_i128 ret, TCGTemp *addr, TCGv_i128 val, + TCGArg idx, MemOp memop, bool new_val, + void (*gen)(TCGv_i64, TCGv_i64, TCGv_i64)) +{ + TCGv_i128 t = tcg_temp_ebb_new_i128(); + TCGv_i128 r = tcg_temp_ebb_new_i128(); + + tcg_gen_qemu_ld_i128_int(r, addr, idx, memop); + gen(TCGV128_LOW(t), TCGV128_LOW(r), TCGV128_LOW(val)); + gen(TCGV128_HIGH(t), TCGV128_HIGH(r), TCGV128_HIGH(val)); + tcg_gen_qemu_st_i128_int(t, addr, idx, memop); + + tcg_gen_mov_i128(ret, r); + tcg_temp_free_i128(t); + tcg_temp_free_i128(r); +} + +static void do_atomic_op_i128(TCGv_i128 ret, TCGTemp *addr, TCGv_i128 val, + TCGArg idx, MemOp memop, void * const table[]) +{ + gen_atomic_op_i128 gen = table[memop & (MO_SIZE | MO_BSWAP)]; + + if (gen) { + MemOpIdx oi = make_memop_idx(memop & ~MO_SIGN, idx); + TCGv_i64 a64 = maybe_extend_addr64(addr); + gen(ret, tcg_env, a64, val, tcg_constant_i32(oi)); + maybe_free_addr64(a64); + return; + } + + gen_helper_exit_atomic(tcg_env); + /* Produce a result */ + tcg_gen_movi_i64(TCGV128_LOW(ret), 0); + tcg_gen_movi_i64(TCGV128_HIGH(ret), 0); +} + +#define GEN_ATOMIC_HELPER128(NAME, OP, NEW) \ +static void * const table_##NAME[(MO_SIZE | MO_BSWAP) + 1] = { \ + [MO_8] = gen_helper_atomic_##NAME##b, \ + [MO_16 | MO_LE] = gen_helper_atomic_##NAME##w_le, \ + [MO_16 | MO_BE] = gen_helper_atomic_##NAME##w_be, \ + [MO_32 | MO_LE] = gen_helper_atomic_##NAME##l_le, \ + [MO_32 | MO_BE] = gen_helper_atomic_##NAME##l_be, \ + WITH_ATOMIC64([MO_64 | MO_LE] = gen_helper_atomic_##NAME##q_le) \ + WITH_ATOMIC64([MO_64 | MO_BE] = gen_helper_atomic_##NAME##q_be) \ + WITH_ATOMIC128([MO_128 | MO_LE] = gen_helper_atomic_##NAME##o_le) \ + WITH_ATOMIC128([MO_128 | MO_BE] = gen_helper_atomic_##NAME##o_be) \ +}; \ +void tcg_gen_atomic_##NAME##_i32_chk(TCGv_i32 ret, TCGTemp *addr, \ + TCGv_i32 val, TCGArg idx, \ + MemOp memop, TCGType addr_type) \ +{ \ + tcg_debug_assert(addr_type == tcg_ctx->addr_type); \ + tcg_debug_assert((memop & MO_SIZE) <= MO_32); \ + if (tcg_ctx->gen_tb->cflags & CF_PARALLEL) { \ + do_atomic_op_i32(ret, addr, val, idx, memop, table_##NAME); \ + } else { \ + do_nonatomic_op_i32(ret, addr, val, idx, memop, NEW, \ + tcg_gen_##OP##_i32); \ + } \ +} \ +void tcg_gen_atomic_##NAME##_i64_chk(TCGv_i64 ret, TCGTemp *addr, \ + TCGv_i64 val, TCGArg idx, \ + MemOp memop, TCGType addr_type) \ +{ \ + tcg_debug_assert(addr_type == tcg_ctx->addr_type); \ + tcg_debug_assert((memop & MO_SIZE) <= MO_64); \ + if (tcg_ctx->gen_tb->cflags & CF_PARALLEL) { \ + do_atomic_op_i64(ret, addr, val, idx, memop, table_##NAME); \ + } else { \ + do_nonatomic_op_i64(ret, addr, val, idx, memop, NEW, \ + tcg_gen_##OP##_i64); \ + } \ +} \ +void tcg_gen_atomic_##NAME##_i128_chk(TCGv_i128 ret, TCGTemp *addr, \ + TCGv_i128 val, TCGArg idx, \ + MemOp memop, TCGType addr_type) \ +{ \ + tcg_debug_assert(addr_type == tcg_ctx->addr_type); \ + tcg_debug_assert((memop & MO_SIZE) == MO_128); \ + if (tcg_ctx->gen_tb->cflags & CF_PARALLEL) { \ + do_atomic_op_i128(ret, addr, val, idx, memop, table_##NAME); \ + } else { \ + do_nonatomic_op_i128(ret, addr, val, idx, memop, NEW, \ + tcg_gen_##OP##_i64); \ + } \ +} + #define GEN_ATOMIC_HELPER(NAME, OP, NEW) \ static void * const table_##NAME[(MO_SIZE | MO_BSWAP) + 1] = { \ [MO_8] = gen_helper_atomic_##NAME##b, \ @@ -1239,8 +1329,8 @@ void tcg_gen_atomic_##NAME##_i64_chk(TCGv_i64 ret, TCGTemp *addr, \ } GEN_ATOMIC_HELPER(fetch_add, add, 0) -GEN_ATOMIC_HELPER(fetch_and, and, 0) -GEN_ATOMIC_HELPER(fetch_or, or, 0) +GEN_ATOMIC_HELPER128(fetch_and, and, 0) +GEN_ATOMIC_HELPER128(fetch_or, or, 0) GEN_ATOMIC_HELPER(fetch_xor, xor, 0) GEN_ATOMIC_HELPER(fetch_smin, smin, 0) GEN_ATOMIC_HELPER(fetch_umin, umin, 0) @@ -1266,6 +1356,7 @@ static void tcg_gen_mov2_i64(TCGv_i64 r, TCGv_i64 a, TCGv_i64 b) tcg_gen_mov_i64(r, b); } -GEN_ATOMIC_HELPER(xchg, mov2, 0) +GEN_ATOMIC_HELPER128(xchg, mov2, 0) #undef GEN_ATOMIC_HELPER +#undef GEN_ATOMIC_HELPER128 From 905c2c34fe3432879f31ef0aa64ce478e573276b Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Sat, 30 Aug 2025 14:50:06 +1000 Subject: [PATCH 29/32] target/arm: Rename isar_feature_aa64_atomics This is FEAT_LSE -- rename the predicate to match. Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Signed-off-by: Peter Maydell Message-id: 20250830045006.380393-1-richard.henderson@linaro.org Message-id: 20250815122653.701782-6-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- bsd-user/aarch64/target_arch_elf.h | 2 +- linux-user/aarch64/elfload.c | 2 +- target/arm/cpu-features.h | 2 +- target/arm/tcg/translate-a64.c | 24 ++++++++++++------------ 4 files changed, 15 insertions(+), 15 deletions(-) diff --git a/bsd-user/aarch64/target_arch_elf.h b/bsd-user/aarch64/target_arch_elf.h index cc87f475b3..cec254f88b 100644 --- a/bsd-user/aarch64/target_arch_elf.h +++ b/bsd-user/aarch64/target_arch_elf.h @@ -114,7 +114,7 @@ static uint32_t get_elf_hwcap(void) GET_FEATURE_ID(aa64_sm3, ARM_HWCAP_A64_SM3); GET_FEATURE_ID(aa64_sm4, ARM_HWCAP_A64_SM4); GET_FEATURE_ID(aa64_fp16, ARM_HWCAP_A64_FPHP | ARM_HWCAP_A64_ASIMDHP); - GET_FEATURE_ID(aa64_atomics, ARM_HWCAP_A64_ATOMICS); + GET_FEATURE_ID(aa64_lse, ARM_HWCAP_A64_ATOMICS); GET_FEATURE_ID(aa64_rdm, ARM_HWCAP_A64_ASIMDRDM); GET_FEATURE_ID(aa64_dp, ARM_HWCAP_A64_ASIMDDP); GET_FEATURE_ID(aa64_fcma, ARM_HWCAP_A64_FCMA); diff --git a/linux-user/aarch64/elfload.c b/linux-user/aarch64/elfload.c index dd5f34398a..8bf39c4730 100644 --- a/linux-user/aarch64/elfload.c +++ b/linux-user/aarch64/elfload.c @@ -154,7 +154,7 @@ abi_ulong get_elf_hwcap(CPUState *cs) GET_FEATURE_ID(aa64_sm3, ARM_HWCAP_A64_SM3); GET_FEATURE_ID(aa64_sm4, ARM_HWCAP_A64_SM4); GET_FEATURE_ID(aa64_fp16, ARM_HWCAP_A64_FPHP | ARM_HWCAP_A64_ASIMDHP); - GET_FEATURE_ID(aa64_atomics, ARM_HWCAP_A64_ATOMICS); + GET_FEATURE_ID(aa64_lse, ARM_HWCAP_A64_ATOMICS); GET_FEATURE_ID(aa64_lse2, ARM_HWCAP_A64_USCAT); GET_FEATURE_ID(aa64_rdm, ARM_HWCAP_A64_ASIMDRDM); GET_FEATURE_ID(aa64_dp, ARM_HWCAP_A64_ASIMDDP); diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index d48754bcf2..451b37b5b3 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -406,7 +406,7 @@ static inline bool isar_feature_aa64_crc32(const ARMISARegisters *id) return FIELD_EX64_IDREG(id, ID_AA64ISAR0, CRC32) != 0; } -static inline bool isar_feature_aa64_atomics(const ARMISARegisters *id) +static inline bool isar_feature_aa64_lse(const ARMISARegisters *id) { return FIELD_EX64_IDREG(id, ID_AA64ISAR0, ATOMIC) >= 2; } diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index 259aa70a36..0ba537268c 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -3237,7 +3237,7 @@ static bool trans_LDXP(DisasContext *s, arg_stxr *a) static bool trans_CASP(DisasContext *s, arg_CASP *a) { - if (!dc_isar_feature(aa64_atomics, s)) { + if (!dc_isar_feature(aa64_lse, s)) { return false; } if (((a->rt | a->rs) & 1) != 0) { @@ -3250,7 +3250,7 @@ static bool trans_CASP(DisasContext *s, arg_CASP *a) static bool trans_CAS(DisasContext *s, arg_CAS *a) { - if (!dc_isar_feature(aa64_atomics, s)) { + if (!dc_isar_feature(aa64_lse, s)) { return false; } gen_compare_and_swap(s, a->rs, a->rt, a->rn, a->sz); @@ -3743,15 +3743,15 @@ static bool do_atomic_ld(DisasContext *s, arg_atomic *a, AtomicThreeOpFn *fn, return true; } -TRANS_FEAT(LDADD, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_add_i64, 0, false) -TRANS_FEAT(LDCLR, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_and_i64, 0, true) -TRANS_FEAT(LDEOR, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_xor_i64, 0, false) -TRANS_FEAT(LDSET, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_or_i64, 0, false) -TRANS_FEAT(LDSMAX, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_smax_i64, MO_SIGN, false) -TRANS_FEAT(LDSMIN, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_smin_i64, MO_SIGN, false) -TRANS_FEAT(LDUMAX, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_umax_i64, 0, false) -TRANS_FEAT(LDUMIN, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_fetch_umin_i64, 0, false) -TRANS_FEAT(SWP, aa64_atomics, do_atomic_ld, a, tcg_gen_atomic_xchg_i64, 0, false) +TRANS_FEAT(LDADD, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_add_i64, 0, false) +TRANS_FEAT(LDCLR, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_and_i64, 0, true) +TRANS_FEAT(LDEOR, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_xor_i64, 0, false) +TRANS_FEAT(LDSET, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_or_i64, 0, false) +TRANS_FEAT(LDSMAX, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_smax_i64, MO_SIGN, false) +TRANS_FEAT(LDSMIN, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_smin_i64, MO_SIGN, false) +TRANS_FEAT(LDUMAX, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_umax_i64, 0, false) +TRANS_FEAT(LDUMIN, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_umin_i64, 0, false) +TRANS_FEAT(SWP, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_xchg_i64, 0, false) static bool trans_LDAPR(DisasContext *s, arg_LDAPR *a) { @@ -3759,7 +3759,7 @@ static bool trans_LDAPR(DisasContext *s, arg_LDAPR *a) TCGv_i64 clean_addr; MemOp mop; - if (!dc_isar_feature(aa64_atomics, s) || + if (!dc_isar_feature(aa64_lse, s) || !dc_isar_feature(aa64_rcpc_8_3, s)) { return false; } From 99e441107efb34a8af990cedbfbda0567d0bb387 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:52 +1000 Subject: [PATCH 30/32] target/arm: Implement FEAT_LSE128 This feature contains the LDCLRP, LDSETP, and SWPP instructions. Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-7-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- target/arm/cpu-features.h | 5 ++++ target/arm/tcg/a64.decode | 7 +++++ target/arm/tcg/translate-a64.c | 49 ++++++++++++++++++++++++++++++++++ 3 files changed, 61 insertions(+) diff --git a/target/arm/cpu-features.h b/target/arm/cpu-features.h index 451b37b5b3..e49e0ae3af 100644 --- a/target/arm/cpu-features.h +++ b/target/arm/cpu-features.h @@ -411,6 +411,11 @@ static inline bool isar_feature_aa64_lse(const ARMISARegisters *id) return FIELD_EX64_IDREG(id, ID_AA64ISAR0, ATOMIC) >= 2; } +static inline bool isar_feature_aa64_lse128(const ARMISARegisters *id) +{ + return FIELD_EX64_IDREG(id, ID_AA64ISAR0, ATOMIC) >= 3; +} + static inline bool isar_feature_aa64_rdm(const ARMISARegisters *id) { return FIELD_EX64_IDREG(id, ID_AA64ISAR0, RDM) != 0; diff --git a/target/arm/tcg/a64.decode b/target/arm/tcg/a64.decode index 766c610c01..55ff6c504f 100644 --- a/target/arm/tcg/a64.decode +++ b/target/arm/tcg/a64.decode @@ -546,6 +546,13 @@ SWP .. 111 0 00 . . 1 ..... 1000 00 ..... ..... @atomic LDAPR sz:2 111 0 00 1 0 1 11111 1100 00 rn:5 rt:5 +# Atomic 128-bit memory operations +&atomic128 rn rt rt2 a r +@atomic128 ........ a:1 r:1 . rt2:5 ...... rn:5 rt:5 &atomic128 +LDCLRP 00011001 . . 1 ..... 000100 ..... ..... @atomic128 +LDSETP 00011001 . . 1 ..... 001100 ..... ..... @atomic128 +SWPP 00011001 . . 1 ..... 100000 ..... ..... @atomic128 + # Load/store register (pointer authentication) # LDRA immediate is 10 bits signed and scaled, but the bits aren't all contiguous diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index 0ba537268c..37bedc3780 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -3753,6 +3753,55 @@ TRANS_FEAT(LDUMAX, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_umax_i64, 0, TRANS_FEAT(LDUMIN, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_fetch_umin_i64, 0, false) TRANS_FEAT(SWP, aa64_lse, do_atomic_ld, a, tcg_gen_atomic_xchg_i64, 0, false) +typedef void Atomic128ThreeOpFn(TCGv_i128, TCGv_i64, TCGv_i128, TCGArg, MemOp); + +static bool do_atomic128_ld(DisasContext *s, arg_atomic128 *a, + Atomic128ThreeOpFn *fn, bool invert) +{ + MemOp mop; + int rlo, rhi; + TCGv_i64 clean_addr, tlo, thi; + TCGv_i128 t16; + + if (a->rt == 31 || a->rt2 == 31 || a->rt == a->rt2) { + return false; + } + if (a->rn == 31) { + gen_check_sp_alignment(s); + } + mop = check_atomic_align(s, a->rn, MO_128); + clean_addr = gen_mte_check1(s, cpu_reg_sp(s, a->rn), false, + a->rn != 31, mop); + + rlo = (s->be_data == MO_LE ? a->rt : a->rt2); + rhi = (s->be_data == MO_LE ? a->rt2 : a->rt); + + tlo = read_cpu_reg(s, rlo, true); + thi = read_cpu_reg(s, rhi, true); + if (invert) { + tcg_gen_not_i64(tlo, tlo); + tcg_gen_not_i64(thi, thi); + } + /* + * The tcg atomic primitives are all full barriers. Therefore we + * can ignore the Acquire and Release bits of this instruction. + */ + t16 = tcg_temp_new_i128(); + tcg_gen_concat_i64_i128(t16, tlo, thi); + + fn(t16, clean_addr, t16, get_mem_index(s), mop); + + tcg_gen_extr_i128_i64(cpu_reg(s, rlo), cpu_reg(s, rhi), t16); + return true; +} + +TRANS_FEAT(LDCLRP, aa64_lse128, do_atomic128_ld, + a, tcg_gen_atomic_fetch_and_i128, true) +TRANS_FEAT(LDSETP, aa64_lse128, do_atomic128_ld, + a, tcg_gen_atomic_fetch_or_i128, false) +TRANS_FEAT(SWPP, aa64_lse128, do_atomic128_ld, + a, tcg_gen_atomic_xchg_i128, false) + static bool trans_LDAPR(DisasContext *s, arg_LDAPR *a) { bool iss_sf = ldst_iss_sf(a->sz, false, false); From 23f5b02447503d9e28f6fad690ccce850c7bb656 Mon Sep 17 00:00:00 2001 From: Richard Henderson Date: Fri, 15 Aug 2025 22:26:53 +1000 Subject: [PATCH 31/32] target/arm: Enable FEAT_LSE128 for -cpu max Signed-off-by: Richard Henderson Reviewed-by: Peter Maydell Message-id: 20250815122653.701782-8-richard.henderson@linaro.org Signed-off-by: Peter Maydell --- docs/system/arm/emulation.rst | 1 + linux-user/aarch64/elfload.c | 1 + target/arm/tcg/cpu64.c | 2 +- 3 files changed, 3 insertions(+), 1 deletion(-) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index b12f013b4f..4e8aca8b5d 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -89,6 +89,7 @@ the following architecture extensions: - FEAT_LRCPC2 (Load-acquire RCpc instructions v2) - FEAT_LSE (Large System Extensions) - FEAT_LSE2 (Large System Extensions v2) +- FEAT_LSE128 (128-bit Atomics) - FEAT_LVA (Large Virtual Address space) - FEAT_MixedEnd (Mixed-endian support) - FEAT_MixedEndEL0 (Mixed-endian support at EL0) diff --git a/linux-user/aarch64/elfload.c b/linux-user/aarch64/elfload.c index 8bf39c4730..77d03b50e1 100644 --- a/linux-user/aarch64/elfload.c +++ b/linux-user/aarch64/elfload.c @@ -216,6 +216,7 @@ abi_ulong get_elf_hwcap2(CPUState *cs) GET_FEATURE_ID(aa64_sme_f16f16, ARM_HWCAP2_A64_SME_F16F16); GET_FEATURE_ID(aa64_sve_b16b16, ARM_HWCAP2_A64_SVE_B16B16); GET_FEATURE_ID(aa64_cssc, ARM_HWCAP2_A64_CSSC); + GET_FEATURE_ID(aa64_lse128, ARM_HWCAP2_A64_LSE128); return hwcaps; } diff --git a/target/arm/tcg/cpu64.c b/target/arm/tcg/cpu64.c index eaf8846a6a..b8b1981e70 100644 --- a/target/arm/tcg/cpu64.c +++ b/target/arm/tcg/cpu64.c @@ -1145,7 +1145,7 @@ void aarch64_max_tcg_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64ISAR0, SHA1, 1); /* FEAT_SHA1 */ t = FIELD_DP64(t, ID_AA64ISAR0, SHA2, 2); /* FEAT_SHA512 */ t = FIELD_DP64(t, ID_AA64ISAR0, CRC32, 1); /* FEAT_CRC32 */ - t = FIELD_DP64(t, ID_AA64ISAR0, ATOMIC, 2); /* FEAT_LSE */ + t = FIELD_DP64(t, ID_AA64ISAR0, ATOMIC, 3); /* FEAT_LSE, FEAT_LSE128 */ t = FIELD_DP64(t, ID_AA64ISAR0, RDM, 1); /* FEAT_RDM */ t = FIELD_DP64(t, ID_AA64ISAR0, SHA3, 1); /* FEAT_SHA3 */ t = FIELD_DP64(t, ID_AA64ISAR0, SM3, 1); /* FEAT_SM3 */ From 2e27650bddd35477d994a795a3b1cb57c8ed5c76 Mon Sep 17 00:00:00 2001 From: Peter Maydell Date: Thu, 21 Aug 2025 16:42:29 +0100 Subject: [PATCH 32/32] hw/arm/stm32f205_soc: Don't leak TYPE_OR_IRQ objects MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit In stm32f250_soc_initfn() we mostly use the standard pattern for child objects of calling object_initialize_child(). However for s->adc_irqs we call object_new() and then later qdev_realize(), and we never unref the object on deinit. This causes a leak, detected by ASAN on the device-introspect-test: Indirect leak of 10 byte(s) in 1 object(s) allocated from: #0 0x5b9fc4789de3 in malloc (/mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/qemu-system-arm+0x21f1de3) (BuildId: 267a2619a026ed91c78a07b1eb2ef15381538efe) #1 0x740de3f28b09 in g_malloc (/lib/x86_64-linux-gnu/libglib-2.0.so.0+0x62b09) (BuildId: 1eb6131419edb83b2178b682829a6913cf682d75) #2 0x740de3f3e4d8 in g_strdup (/lib/x86_64-linux-gnu/libglib-2.0.so.0+0x784d8) (BuildId: 1eb6131419edb83b2178b682829a6913cf682d75) #3 0x5b9fc70159e1 in g_strdup_inline /usr/include/glib-2.0/glib/gstrfuncs.h:321:10 #4 0x5b9fc70159e1 in object_property_try_add /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:1276:18 #5 0x5b9fc7015f94 in object_property_add /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:1294:12 #6 0x5b9fc701b900 in object_add_link_prop /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:2021:10 #7 0x5b9fc701b3fc in object_property_add_link /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:2037:12 #8 0x5b9fc4c299fb in qdev_init_gpio_out_named /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../hw/core/gpio.c:90:9 #9 0x5b9fc4c29b26 in qdev_init_gpio_out /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../hw/core/gpio.c:101:5 #10 0x5b9fc4c0f77a in or_irq_init /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../hw/core/or-irq.c:70:5 #11 0x5b9fc70257e1 in object_init_with_type /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:428:9 #12 0x5b9fc700cd4b in object_initialize_with_type /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:570:5 #13 0x5b9fc700e66d in object_new_with_type /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:774:5 #14 0x5b9fc700e750 in object_new /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../qom/object.c:789:12 #15 0x5b9fc68b2162 in stm32f205_soc_initfn /mnt/nvmedisk/linaro/qemu-from-laptop/qemu/build/arm-asan/../../hw/arm/stm32f205_soc.c:69:26 Switch to using object_initialize_child() like all our other child objects for this SoC object. Cc: qemu-stable@nongnu.org Fixes: b63041c8f6b ("STM32F205: Connect the ADC devices") Signed-off-by: Peter Maydell Reviewed-by: Philippe Mathieu-Daudé Message-id: 20250821154229.2417453-1-peter.maydell@linaro.org --- hw/arm/stm32f205_soc.c | 10 +++++----- include/hw/arm/stm32f205_soc.h | 2 +- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/hw/arm/stm32f205_soc.c b/hw/arm/stm32f205_soc.c index 229af7fb10..e3c7203c6e 100644 --- a/hw/arm/stm32f205_soc.c +++ b/hw/arm/stm32f205_soc.c @@ -66,7 +66,7 @@ static void stm32f205_soc_initfn(Object *obj) TYPE_STM32F2XX_TIMER); } - s->adc_irqs = OR_IRQ(object_new(TYPE_OR_IRQ)); + object_initialize_child(obj, "adc-irq-orgate", &s->adc_irqs, TYPE_OR_IRQ); for (i = 0; i < STM_NUM_ADCS; i++) { object_initialize_child(obj, "adc[*]", &s->adc[i], TYPE_STM32F2XX_ADC); @@ -171,12 +171,12 @@ static void stm32f205_soc_realize(DeviceState *dev_soc, Error **errp) } /* ADC 1 to 3 */ - object_property_set_int(OBJECT(s->adc_irqs), "num-lines", STM_NUM_ADCS, + object_property_set_int(OBJECT(&s->adc_irqs), "num-lines", STM_NUM_ADCS, &error_abort); - if (!qdev_realize(DEVICE(s->adc_irqs), NULL, errp)) { + if (!qdev_realize(DEVICE(&s->adc_irqs), NULL, errp)) { return; } - qdev_connect_gpio_out(DEVICE(s->adc_irqs), 0, + qdev_connect_gpio_out(DEVICE(&s->adc_irqs), 0, qdev_get_gpio_in(armv7m, ADC_IRQ)); for (i = 0; i < STM_NUM_ADCS; i++) { @@ -187,7 +187,7 @@ static void stm32f205_soc_realize(DeviceState *dev_soc, Error **errp) busdev = SYS_BUS_DEVICE(dev); sysbus_mmio_map(busdev, 0, adc_addr[i]); sysbus_connect_irq(busdev, 0, - qdev_get_gpio_in(DEVICE(s->adc_irqs), i)); + qdev_get_gpio_in(DEVICE(&s->adc_irqs), i)); } /* SPI 1 and 2 */ diff --git a/include/hw/arm/stm32f205_soc.h b/include/hw/arm/stm32f205_soc.h index 4f4c8bbebc..46eda3403a 100644 --- a/include/hw/arm/stm32f205_soc.h +++ b/include/hw/arm/stm32f205_soc.h @@ -59,7 +59,7 @@ struct STM32F205State { STM32F2XXADCState adc[STM_NUM_ADCS]; STM32F2XXSPIState spi[STM_NUM_SPIS]; - OrIRQState *adc_irqs; + OrIRQState adc_irqs; MemoryRegion sram; MemoryRegion flash;