- Linux-kselftest-mirror - lists.linaro.org

[PATCH v2 00/10] x86/module: rework ROX cache to avoid writable copy

by Mike Rapoport

From: "Mike Rapoport (Microsoft)" <rppt(a)kernel.org> Hi, Following Peter's comments [1] these patches rework handling of ROX caches for module text allocations. Instead of using a writable copy that really complicates alternatives patching, temporarily remap parts of a large ROX page as RW for the time of module formation and then restore it's ROX protections when the module is ready. To keep the ROX memory mapped with large pages, make set_memory_rox() capable of restoring large pages (more details are in patch 3). Since this is really about x86, I believe this should go in via tip tree. The patches also available in git https://git.kernel.org/rppt/h/execmem/x86-rox/v9 v2 changes: * only collapse large mappings in set_memory_rox() * simplify RW <-> ROX remapping * don't remove ROX cache pages from the direct map (patch 4) v1: https://lore.kernel.org/all/20241227072825.1288491-1-rppt@kernel.org [1] https://lore.kernel.org/all/20241209083818.GK8562@noisy.programming.kicks-a… Kirill A. Shutemov (1): x86/mm/pat: restore large ROX pages after fragmentation Mike Rapoport (Microsoft) (9): x86/mm/pat: cpa-test: fix length for CPA_ARRAY test x86/mm/pat: drop duplicate variable in cpa_flush() execmem: don't remove ROX cache from the direct map execmem: add API for temporal remapping as RW and restoring ROX afterwards module: introduce MODULE_STATE_GONE module: switch to execmem API for remapping as RW and restoring ROX Revert "x86/module: prepare module loading for ROX allocations of text" module: drop unused module_writable_address() x86: re-enable EXECMEM_ROX support arch/um/kernel/um_arch.c | 11 +- arch/x86/Kconfig | 1 + arch/x86/entry/vdso/vma.c | 3 +- arch/x86/include/asm/alternative.h | 14 +- arch/x86/include/asm/pgtable_types.h | 2 + arch/x86/kernel/alternative.c | 181 ++++++-------- arch/x86/kernel/ftrace.c | 30 ++- arch/x86/kernel/module.c | 45 ++-- arch/x86/mm/pat/cpa-test.c | 2 +- arch/x86/mm/pat/set_memory.c | 220 +++++++++++++++++- include/linux/execmem.h | 31 +++ include/linux/module.h | 22 +- include/linux/moduleloader.h | 4 - include/linux/vm_event_item.h | 2 + kernel/module/kallsyms.c | 8 +- kernel/module/kdb.c | 2 +- kernel/module/main.c | 86 ++----- kernel/module/procfs.c | 2 +- kernel/module/strict_rwx.c | 9 +- kernel/tracepoint.c | 2 + lib/kunit/test.c | 2 + mm/execmem.c | 39 ++-- mm/vmstat.c | 2 + samples/livepatch/livepatch-callbacks-demo.c | 1 + .../test_modules/test_klp_callbacks_demo.c | 1 + .../test_modules/test_klp_callbacks_demo2.c | 1 + .../livepatch/test_modules/test_klp_state.c | 1 + .../livepatch/test_modules/test_klp_state2.c | 1 + 28 files changed, 442 insertions(+), 283 deletions(-) base-commit: ffd294d346d185b70e28b1a28abe367bbfe53c04 -- 2.45.2

5 months, 1 week

4
16
0 0

[PATCH net v2] bonding: Correctly support GSO ESP offload

by Cosmin Ratiu

The referenced fix is incomplete. It correctly computes bond_dev->gso_partial_features across slaves, but unfortunately netdev_fix_features discards gso_partial_features from the feature set if NETIF_F_GSO_PARTIAL isn't set in bond_dev->features. This is visible with ethtool -k bond0 | grep esp: tx-esp-segmentation: off [requested on] esp-hw-offload: on esp-tx-csum-hw-offload: on This patch reworks the bonding GSO offload support by: - making aggregating gso_partial_features across slaves similar to the other feature sets (this part is a no-op). - advertising the default partial gso features on empty bond devs, same as with other feature sets (also a no-op). - adding NETIF_F_GSO_PARTIAL to hw_enc_features filtered across slaves. - adding NETIF_F_GSO_PARTIAL to features in bond_setup() With all of these, 'ethtool -k bond0 | grep esp' now reports: tx-esp-segmentation: on esp-hw-offload: on esp-tx-csum-hw-offload: on Fixes: 4861333b4217 ("bonding: add ESP offload features when slaves support") Signed-off-by: Cosmin Ratiu <cratiu(a)nvidia.com> --- drivers/net/bonding/bond_main.c | 19 ++++++++++--------- 1 file changed, 10 insertions(+), 9 deletions(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 7b78c2bada81..e45bba240cbc 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1538,17 +1538,20 @@ static netdev_features_t bond_fix_features(struct net_device *dev, NETIF_F_HIGHDMA | NETIF_F_LRO) #define BOND_ENC_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \ - NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE) + NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE | \ + NETIF_F_GSO_PARTIAL) #define BOND_MPLS_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \ NETIF_F_GSO_SOFTWARE) +#define BOND_GSO_PARTIAL_FEATURES (NETIF_F_GSO_ESP) + static void bond_compute_features(struct bonding *bond) { + netdev_features_t gso_partial_features = BOND_GSO_PARTIAL_FEATURES; unsigned int dst_release_flag = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; - netdev_features_t gso_partial_features = NETIF_F_GSO_ESP; netdev_features_t vlan_features = BOND_VLAN_FEATURES; netdev_features_t enc_features = BOND_ENC_FEATURES; #ifdef CONFIG_XFRM_OFFLOAD @@ -1582,8 +1585,9 @@ static void bond_compute_features(struct bonding *bond) BOND_XFRM_FEATURES); #endif /* CONFIG_XFRM_OFFLOAD */ - if (slave->dev->hw_enc_features & NETIF_F_GSO_PARTIAL) - gso_partial_features &= slave->dev->gso_partial_features; + gso_partial_features = netdev_increment_features(gso_partial_features, + slave->dev->gso_partial_features, + BOND_GSO_PARTIAL_FEATURES); mpls_features = netdev_increment_features(mpls_features, slave->dev->mpls_features, @@ -1598,12 +1602,8 @@ static void bond_compute_features(struct bonding *bond) } bond_dev->hard_header_len = max_hard_header_len; - if (gso_partial_features & NETIF_F_GSO_ESP) - bond_dev->gso_partial_features |= NETIF_F_GSO_ESP; - else - bond_dev->gso_partial_features &= ~NETIF_F_GSO_ESP; - done: + bond_dev->gso_partial_features = gso_partial_features; bond_dev->vlan_features = vlan_features; bond_dev->hw_enc_features = enc_features | NETIF_F_GSO_ENCAP_ALL | NETIF_F_HW_VLAN_CTAG_TX | @@ -6046,6 +6046,7 @@ void bond_setup(struct net_device *bond_dev) bond_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL; bond_dev->features |= bond_dev->hw_features; bond_dev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX; + bond_dev->features |= NETIF_F_GSO_PARTIAL; #ifdef CONFIG_XFRM_OFFLOAD bond_dev->hw_features |= BOND_XFRM_FEATURES; /* Only enable XFRM features if this is an active-backup config */ -- 2.45.0

5 months, 1 week

2
3
0 0

[PATCH bpf-next v2 00/10] selftests/bpf: Migrate test_xdp_redirect_multi.sh to test_progs

by Bastien Curutchet (eBPF Foundation)

Hi all, This patch series continues the work to migrate the *.sh tests into prog_tests framework. test_xdp_redirect_multi.sh tests the XDP redirections done through bpf_redirect_map(). This is already partly covered by test_xdp_veth.c that already tests map redirections at XDP level. What isn't covered yet by test_xdp_veth is the use of the broadcast flags (BPF_F_BROADCAST or BPF_F_EXCLUDE_INGRESS) and XDP egress programs. Hence, this patch series add test cases to test_xdp_veth.c to get rid of the test_xdp_redirect_multi.sh: - PATCH 1 to 5 rework test_xdp_veth to make it more generic and allow to configure different test cases - PATCH 6 adds test cases for 'classic' bpf_redirect_map() - PATCH 7 & 8 covers the broadcast flags - PATCH 9 covers the XDP egress programs - PATCH 10 removes test_xdp_redirect_multi.sh Signed-off-by: Bastien Curutchet (eBPF Foundation) <bastien.curutchet(a)bootlin.com> --- Changes in v2: - Use serial_test_* to avoid conflict between tests - Link to v1: https://lore.kernel.org/r/20250121-redirect-multi-v1-0-b215e35ff505@bootlin… --- Bastien Curutchet (eBPF Foundation) (10): selftests/bpf: test_xdp_veth: Split network configuration selftests/bpf: Remove unused argument selftests/bpf: test_xdp_veth: Rename config[] selftests/bpf: test_xdp_veth: Add prog_config[] table selftests/bpf: test_xdp_veth: Add XDP flags to prog_configuration selftests/bpf: test_xdp_veth: Add new test cases for XDP flags selftests/bpf: Optionally select broadcasting flags selftests/bpf: test_xdp_veth: Add XDP broadcast redirection tests selftests/bpf: test_xdp_veth: Add XDP program on egress test selftests/bpf: Remove test_xdp_redirect_multi.sh tools/testing/selftests/bpf/Makefile | 2 - .../selftests/bpf/prog_tests/test_xdp_veth.c | 534 +++++++++++++++++---- .../testing/selftests/bpf/progs/xdp_redirect_map.c | 89 ++++ .../selftests/bpf/progs/xdp_redirect_multi_kern.c | 41 +- .../selftests/bpf/test_xdp_redirect_multi.sh | 214 --------- tools/testing/selftests/bpf/xdp_redirect_multi.c | 226 --------- 6 files changed, 553 insertions(+), 553 deletions(-) --- base-commit: 349e0551b929b4712b4d6127f67dfa41ed48d5a2 change-id: 20250103-redirect-multi-245d6eafb5d1 Best regards, -- Bastien Curutchet (eBPF Foundation) <bastien.curutchet(a)bootlin.com>

5 months, 1 week

3
12
0 0

[PATCH v1 0/2] mm: Clear uffd-wp PTE/PMD state on mremap()

by Ryan Roberts

Hi All, This series contains a fix for a warning emitted when a uffd-registered region, which doesn't have UFFD_FEATURE_EVENT_REMAP, is mremap()ed. patch 1 describes the problem and fixes it, and patch 2 adds a selftest to verify the fix. Thanks to Mikołaj Lenczewski who originally created the patch, which I have subsequently extended. Applies on top of mm-unstable (f349e79bfbf3) Thanks, Ryan Ryan Roberts (2): mm: Clear uffd-wp PTE/PMD state on mremap() selftests/mm: Introduce uffd-wp-mremap regression test include/linux/userfaultfd_k.h | 12 + mm/huge_memory.c | 12 + mm/hugetlb.c | 14 +- mm/mremap.c | 32 +- tools/testing/selftests/mm/.gitignore | 1 + tools/testing/selftests/mm/Makefile | 2 + tools/testing/selftests/mm/run_vmtests.sh | 1 + tools/testing/selftests/mm/uffd-wp-mremap.c | 380 ++++++++++++++++++++ 8 files changed, 452 insertions(+), 2 deletions(-) create mode 100644 tools/testing/selftests/mm/uffd-wp-mremap.c -- 2.43.0

5 months, 1 week

5
14
0 0

[PATCH net v2] bonding: Correctly support GSO ESP offload

by Cosmin Ratiu

V2: Removed unnecessary Change-Id Changed "bond->features" to "bond_dev->features" in commit message.

5 months, 1 week

1
0
0 0

[PATCH net] bonding: Correctly support GSO ESP offload

by Cosmin Ratiu

The referenced fix is incomplete. It correctly computes bond_dev->gso_partial_features across slaves, but unfortunately netdev_fix_features discards gso_partial_features from the feature set if NETIF_F_GSO_PARTIAL isn't set in bond->features. This is visible with ethtool -k bond0 | grep esp: tx-esp-segmentation: off [requested on] esp-hw-offload: on esp-tx-csum-hw-offload: on This patch reworks the bonding GSO offload support by: - making aggregating gso_partial_features across slaves similar to the other feature sets (this part is a no-op). - adding NETIF_F_GSO_PARTIAL to hw_enc_features filtered across slaves. - adding NETIF_F_GSO_PARTIAL to features in bond_setup() With all of these, 'ethtool -k bond0 | grep esp' now reports: tx-esp-segmentation: on esp-hw-offload: on esp-tx-csum-hw-offload: on Fixes: 4861333b4217 ("bonding: add ESP offload features when slaves support") Signed-off-by: Cosmin Ratiu <cratiu(a)nvidia.com> Change-Id: Iebd2a9d903d3e056e7717e8ca2527a9adf21b2e1 --- drivers/net/bonding/bond_main.c | 18 ++++++++++-------- 1 file changed, 10 insertions(+), 8 deletions(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 7b78c2bada81..de105868c009 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -1538,17 +1538,20 @@ static netdev_features_t bond_fix_features(struct net_device *dev, NETIF_F_HIGHDMA | NETIF_F_LRO) #define BOND_ENC_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \ - NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE) + NETIF_F_RXCSUM | NETIF_F_GSO_SOFTWARE | \ + NETIF_F_GSO_PARTIAL) #define BOND_MPLS_FEATURES (NETIF_F_HW_CSUM | NETIF_F_SG | \ NETIF_F_GSO_SOFTWARE) +#define BOND_GSO_PARTIAL_FEATURES (NETIF_F_GSO_ESP) + static void bond_compute_features(struct bonding *bond) { + netdev_features_t gso_partial_features = BOND_GSO_PARTIAL_FEATURES; unsigned int dst_release_flag = IFF_XMIT_DST_RELEASE | IFF_XMIT_DST_RELEASE_PERM; - netdev_features_t gso_partial_features = NETIF_F_GSO_ESP; netdev_features_t vlan_features = BOND_VLAN_FEATURES; netdev_features_t enc_features = BOND_ENC_FEATURES; #ifdef CONFIG_XFRM_OFFLOAD @@ -1582,8 +1585,9 @@ static void bond_compute_features(struct bonding *bond) BOND_XFRM_FEATURES); #endif /* CONFIG_XFRM_OFFLOAD */ - if (slave->dev->hw_enc_features & NETIF_F_GSO_PARTIAL) - gso_partial_features &= slave->dev->gso_partial_features; + gso_partial_features = netdev_increment_features(gso_partial_features, + slave->dev->gso_partial_features, + BOND_GSO_PARTIAL_FEATURES); mpls_features = netdev_increment_features(mpls_features, slave->dev->mpls_features, @@ -1598,10 +1602,7 @@ static void bond_compute_features(struct bonding *bond) } bond_dev->hard_header_len = max_hard_header_len; - if (gso_partial_features & NETIF_F_GSO_ESP) - bond_dev->gso_partial_features |= NETIF_F_GSO_ESP; - else - bond_dev->gso_partial_features &= ~NETIF_F_GSO_ESP; + bond_dev->gso_partial_features = gso_partial_features; done: bond_dev->vlan_features = vlan_features; @@ -6046,6 +6047,7 @@ void bond_setup(struct net_device *bond_dev) bond_dev->hw_features |= NETIF_F_GSO_ENCAP_ALL; bond_dev->features |= bond_dev->hw_features; bond_dev->features |= NETIF_F_HW_VLAN_CTAG_TX | NETIF_F_HW_VLAN_STAG_TX; + bond_dev->features |= NETIF_F_GSO_PARTIAL; #ifdef CONFIG_XFRM_OFFLOAD bond_dev->hw_features |= BOND_XFRM_FEATURES; /* Only enable XFRM features if this is an active-backup config */ -- 2.45.0

5 months, 1 week

2
2
0 0

[PATCH 0/3] tools/nolibc: compatibility with -Wmissing-prototypes

by Thomas Weißschuh

Make nolibc compatible with -Wmissing-prototypes Signed-off-by: Thomas Weißschuh <linux(a)weissschuh.net> --- Thomas Weißschuh (3): tools/nolibc: add prototypes for non-static functions selftests/nolibc: ignore -Wmissing-prototypes selftests/nolibc: enable -Wmissing-prototypes tools/include/nolibc/arch-mips.h | 1 + tools/include/nolibc/crt.h | 2 ++ tools/include/nolibc/signal.h | 1 + tools/include/nolibc/stackprotector.h | 2 ++ tools/include/nolibc/stdlib.h | 1 + tools/include/nolibc/string.h | 4 ++++ tools/testing/selftests/nolibc/Makefile | 2 +- tools/testing/selftests/nolibc/nolibc-test.c | 2 ++ 8 files changed, 14 insertions(+), 1 deletion(-) --- base-commit: d0d106a2bd21499901299160744e5fe9f4c83ddb change-id: 20250123-nolibc-prototype-0477bb05c808 Best regards, -- Thomas Weißschuh <linux(a)weissschuh.net>

5 months, 1 week

2
4
0 0

[PATCH v6 0/3] Enable Zicbom in usermode

by Yunhui Cui

v1/v2: There is only the first patch: RISC-V: Enable cbo.clean/flush in usermode, which mainly removes the enabling of cbo.inval in user mode. v3: Add the functionality of Expose Zicbom and selftests for Zicbom. v4: Modify the order of macros, The test_no_cbo_inval function is added separately. v5: 1. Modify the order of RISCV_HWPROBE_KEY_ZICBOM_BLOCK_SIZE in hwprobe.rst 2. "TEST_NO_ZICBOINVAL" -> "TEST_NO_CBO_INVAL" v6: Change hwprobe_ext0_has's second param to u64. Yunhui Cui (3): RISC-V: Enable cbo.clean/flush in usermode RISC-V: hwprobe: Expose Zicbom extension and its block size RISC-V: selftests: Add TEST_ZICBOM into CBO tests Documentation/arch/riscv/hwprobe.rst | 6 ++ arch/riscv/include/asm/hwprobe.h | 2 +- arch/riscv/include/uapi/asm/hwprobe.h | 2 + arch/riscv/kernel/cpufeature.c | 8 +++ arch/riscv/kernel/sys_hwprobe.c | 8 ++- tools/testing/selftests/riscv/hwprobe/cbo.c | 66 +++++++++++++++++---- 6 files changed, 79 insertions(+), 13 deletions(-) -- 2.39.2

5 months, 1 week

1
3
0 0

[PATCH RFC 5/5] rcutorture: kvm: Invoke stress-ng building it if necessary

by Joel Fernandes (Google)

Invoke stress-ng from init using new --stress-ng and optional --stress-ng-args. Default --stress-ng-args are used if none are provided. Not passing --stress-ng does not change the behavior of kvm.sh from before. Signed-off-by: Joel Fernandes (Google) <joel(a)joelfernandes.org> --- tools/testing/selftests/rcutorture/bin/kvm.sh | 38 +++++++++++++++++-- 1 file changed, 34 insertions(+), 4 deletions(-) diff --git a/tools/testing/selftests/rcutorture/bin/kvm.sh b/tools/testing/selftests/rcutorture/bin/kvm.sh index 4766c3023fed..d35496247ee6 100755 --- a/tools/testing/selftests/rcutorture/bin/kvm.sh +++ b/tools/testing/selftests/rcutorture/bin/kvm.sh @@ -42,13 +42,15 @@ TORTURE_JITTER_STOP="" TORTURE_KCONFIG_KASAN_ARG="" TORTURE_KCONFIG_KCSAN_ARG="" TORTURE_KMAKE_ARG="" +TORTURE_MOD=rcutorture TORTURE_NO_AFFINITY="" TORTURE_QEMU_MEM=512 torture_qemu_mem_default=1 TORTURE_REMOTE= TORTURE_SHUTDOWN_GRACE=180 +TORTURE_STRESS_NG= +TORTURE_STRESS_NG_DEFAULT_ARGS="--cpu 1 --cpu-method matrixprod --cpu-ops 1000000 --perf -t 5" TORTURE_SUITE=rcu -TORTURE_MOD=rcutorture TORTURE_TRUST_MAKE="" debuginfo="CONFIG_DEBUG_INFO_NONE=n CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y" resdir="" @@ -90,6 +92,8 @@ usage () { echo " --remote" echo " --results absolute-pathname" echo " --shutdown-grace seconds" + echo " --stress-ng" + echo " --stress-ng-args \"stress-ng arguments\"" echo " --torture lock|rcu|rcuscale|refscale|scf|X*" echo " --trust-make" exit 1 @@ -251,6 +255,14 @@ do TORTURE_SHUTDOWN_GRACE=$2 shift ;; + --stress-ng) + TORTURE_STRESS_NG=1 + ;; + --stress-ng-args) + checkarg --stress-ng-args "(stress-ng arguments)" "$#" "$2" '.*' '^error' + TORTURE_STRESS_NG_DEFAULT_ARGS="$2" + shift + ;; --torture) checkarg --torture "(suite name)" "$#" "$2" '^$lock\|rcu\|rcuscale\|refscale\|scf\|X.*$$' '^--' TORTURE_SUITE=$2 @@ -275,9 +287,27 @@ do shift done -if test -z "$dryrun" && test -n "$TORTURE_INITRD" && !tools/testing/selftests/rcutorture/bin/mkinitrd.sh - echo No initrd and unable to create one, aborting test >&2 - exit 1 +if test -n "$TORTURE_STRESS_NG" +then + if ! "$RCUTORTURE/bin/mkstress-ng.sh" + then + echo "Failed to build stress-ng, aborting test" >&2 + exit 1 + fi +fi + +if test -z "$dryrun" && test -n "$TORTURE_INITRD" +then + stress_args="" + if test -n "$TORTURE_STRESS_NG" + then + stress_args="stress-ng $TORTURE_STRESS_NG_DEFAULT_ARGS" + fi + if ! "$RCUTORTURE/bin/mkinitrd.sh" $stress_args + then + echo "No initrd and unable to create one, aborting test" >&2 + exit 1 + fi fi CONFIGFRAG=${RCUTORTURE}/configs/${TORTURE_SUITE}; export CONFIGFRAG -- 2.34.1

5 months, 1 week

1
0
0 0

[PATCH RFC 4/5] rcutorture: mkinitrd: Use previous init.c to check if rebuild needed

by Joel Fernandes (Google)

Create the init.c as a temporary init-tmp.c file, even if init already exists. Ensure the existing init.c matches this. If init.c doesn't exist in initrd/ then skip the check and assume init-tmp.c as the new init.c. After that, check if init binary already exists as we did before and skip the build if it does. This does mean we keep a copy of init.c around in the initrd but this is a small file and it is useful to see for debugging anyway. Signed-off-by: Joel Fernandes (Google) <joel(a)joelfernandes.org> --- .../selftests/rcutorture/bin/mkinitrd.sh | 34 +++++++++++++------ 1 file changed, 24 insertions(+), 10 deletions(-) diff --git a/tools/testing/selftests/rcutorture/bin/mkinitrd.sh b/tools/testing/selftests/rcutorture/bin/mkinitrd.sh index 4ba5e962e3cf..d9fbfa205384 100755 --- a/tools/testing/selftests/rcutorture/bin/mkinitrd.sh +++ b/tools/testing/selftests/rcutorture/bin/mkinitrd.sh @@ -17,10 +17,6 @@ if [ ! -d "$D" ]; then echo >&2 "$D does not exist: Malformed kernel source tree?" exit 1 fi -if [ -s "$D/initrd/init" ]; then - echo "$D/initrd/init already exists, no need to create it" - exit 0 -fi # Create a C-language initrd/init infinite-loop program and statically # link it. This results in a very small initrd. @@ -29,8 +25,10 @@ cd $D mkdir -p initrd cd initrd -# Generate the init.c with optional command -cat > init.c << 'EOF_HEAD' +# Generate an init-tmp.c with optional command. This will then be compared +# with any existing init.c. The reason for this is, we want to force a +# rebuild if the optional command or command line arguments have changed. +cat > init-tmp.c << 'EOF_HEAD' #ifndef NOLIBC #include <unistd.h> #include <sys/time.h> @@ -45,7 +43,7 @@ if [ $# -gt 0 ]; then # If command provided, generate run_optional_command() with the specified command. # We use printf to generate the command and args. # Example: echo $(printf '"%s", ' cmd a1 a2) gives: "cmd", "a1", "a2", - cat >> init.c << EOF + cat >> init-tmp.c << EOF pid_t pid = fork(); if (pid == 0) { char *args[] = {$(printf '"%s", ' "$@")NULL}; @@ -54,11 +52,11 @@ if [ $# -gt 0 ]; then EOF else # If no command provided, function will be empty - echo " /* No command specified */" >> init.c + echo " /* No command specified */" >> init-tmp.c fi # Add the rest of the program -cat >> init.c << 'EOF_TAIL' +cat >> init-tmp.c << 'EOF_TAIL' } int main(int argc, char *argv[]) @@ -95,6 +93,23 @@ int main(int argc, char *argv[]) } EOF_TAIL +# Check if init.c exists and compare with init-tmp.c +if [ -f "init.c" ]; then + if ! cmp -s "init.c" "init-tmp.c"; then + mv "init-tmp.c" "init.c" + else + rm "init-tmp.c" + fi +else + mv "init-tmp.c" "init.c" +fi + +# Now check if init binary exists and is up to date +if [ -s "init" ] && [ "init" -nt "init.c" ]; then + echo "$D/initrd/init already exists and is up to date" + exit 0 +fi + # build using nolibc on supported archs (smaller executable) and fall # back to regular glibc on other ones. if echo -e "#if __x86_64__||__i386__||__i486__||__i586__||__i686__" \ @@ -120,7 +135,6 @@ then exit "$ret" fi -rm init.c echo "Done creating a statically linked C-language initrd" exit 0 -- 2.34.1

5 months, 1 week

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror