A few archs like powerpc have different errno.h values for macros EDEADLOCK and EDEADLK. In code including both libc and linux versions of errno.h, this can result in multiple definitions of EDEADLOCK in the include chain. Definitions to the same value (e.g. seen with mips) do not raise warnings, but on powerpc there are redefinitions changing the value, which raise warnings and errors (if using "-Werror").
Guard against these redefinitions to avoid build errors like the following, first seen cross-compiling libbpf v5.8.9 for powerpc using GCC 8.4.0 with musl 1.1.24:
In file included from ../../arch/powerpc/include/uapi/asm/errno.h:5, from ../../include/linux/err.h:8, from libbpf.c:29: ../../include/uapi/asm-generic/errno.h:40: error: "EDEADLOCK" redefined [-Werror] #define EDEADLOCK EDEADLK
In file included from toolchain-powerpc_8540_gcc-8.4.0_musl/include/errno.h:10, from libbpf.c:26: toolchain-powerpc_8540_gcc-8.4.0_musl/include/bits/errno.h:58: note: this is the location of the previous definition #define EDEADLOCK 58
cc1: all warnings being treated as errors
CC: Stable stable@vger.kernel.org Reported-by: Rosen Penev rosenp@gmail.com Signed-off-by: Tony Ambardar Tony.Ambardar@gmail.com --- v1 -> v2: * clean up commit description formatting
v2 -> v3: (per Michael Ellerman) * drop indeterminate 'Fixes' tags, request stable backports instead --- arch/powerpc/include/uapi/asm/errno.h | 1 + tools/arch/powerpc/include/uapi/asm/errno.h | 1 + 2 files changed, 2 insertions(+)
diff --git a/arch/powerpc/include/uapi/asm/errno.h b/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/arch/powerpc/include/uapi/asm/errno.h +++ b/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK diff --git a/tools/arch/powerpc/include/uapi/asm/errno.h b/tools/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/tools/arch/powerpc/include/uapi/asm/errno.h +++ b/tools/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK
Hi
[This is an automated email]
This commit has been processed because it contains a -stable tag. The stable tag indicates that it's relevant for the following trees: all
The bot has tested the following trees: v5.8.10, v5.4.66, v4.19.146, v4.14.198, v4.9.236, v4.4.236.
v5.8.10: Build OK! v5.4.66: Build OK! v4.19.146: Build OK! v4.14.198: Failed to apply! Possible dependencies: 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") a3f22d505f56 ("s390/perf: add callback to perf to enable using AUX buffer") a81c42136604 ("perf s390: add regs_query_register_offset()") a9fc2db0a8ab ("s390/perf: define common DWARF register string table") f704ef44602f ("s390/perf: add support for perf_regs and libdw")
v4.9.236: Failed to apply! Possible dependencies: 0c744ea4f77d ("Linux 4.10-rc2") 2bd6bf03f4c1 ("Linux 4.14-rc1") 2ea659a9ef48 ("Linux 4.12-rc1") 49def1853334 ("Linux 4.10-rc4") 566cf877a1fc ("Linux 4.10-rc6") 5771a8c08880 ("Linux v4.13-rc1") 7089db84e356 ("Linux 4.10-rc8") 7a308bb3016f ("Linux 4.10-rc5") 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 7ce7d89f4883 ("Linux 4.10-rc1") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") a121103c9228 ("Linux 4.10-rc3") a81c42136604 ("perf s390: add regs_query_register_offset()") a9fc2db0a8ab ("s390/perf: define common DWARF register string table") b24413180f56 ("License cleanup: add SPDX GPL-2.0 license identifier to files with no license") c1ae3cfa0e89 ("Linux 4.11-rc1") c470abd4fde4 ("Linux 4.10") d5adbfcd5f7b ("Linux 4.10-rc7")
v4.4.236: Failed to apply! Possible dependencies: 0c4d40d58075 ("tools build: Add BPF feature check to test-all") 1925459b4d92 ("tools build: Fix feature Makefile issues with 'O='") 58683600dfe3 ("perf build: Use FEATURE-DUMP in bpf subproject") 76ee2ff34274 ("tools build feature: Move dwarf post unwind choice output into perf") 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 8ee4646038e4 ("perf build: Add libcrypto feature detection") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") 96b9e70b8e6c ("perf build: Introduce FEATURES_DUMP make variable") 9fd4186ac19a ("tools build: Allow subprojects select all feature checkers") abb26210a395 ("perf tools: Force fixdep compilation at the start of the build") aeafd623f866 ("perf tools: Move headers check into bash script") c053a1506fae ("perf build: Select all feature checkers for feature-dump") d4dfdf00d43e ("perf jvmti: Plug compilation into perf build") d58ac0bf8d1e ("perf build: Add clang and llvm compile and linking support") d8ad6a15cc3a ("tools lib bpf: Don't do a feature check when cleaning") e12b202f8fb9 ("perf jitdump: Build only on supported archs") e26e63be64a1 ("perf build: Add sdt feature detection")
NOTE: The patch will not be queued to stable trees until it is upstream.
How should we proceed with this patch?
On Mon, 21 Sep 2020 at 05:54, Sasha Levin sashal@kernel.org wrote:
Hi
[This is an automated email]
This commit has been processed because it contains a -stable tag. The stable tag indicates that it's relevant for the following trees: all
The bot has tested the following trees: v5.8.10, v5.4.66, v4.19.146, v4.14.198, v4.9.236, v4.4.236.
v5.8.10: Build OK! v5.4.66: Build OK! v4.19.146: Build OK! v4.14.198: Failed to apply! Possible dependencies: 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") a3f22d505f56 ("s390/perf: add callback to perf to enable using AUX buffer") a81c42136604 ("perf s390: add regs_query_register_offset()") a9fc2db0a8ab ("s390/perf: define common DWARF register string table") f704ef44602f ("s390/perf: add support for perf_regs and libdw")
v4.9.236: Failed to apply! Possible dependencies: 0c744ea4f77d ("Linux 4.10-rc2") 2bd6bf03f4c1 ("Linux 4.14-rc1") 2ea659a9ef48 ("Linux 4.12-rc1") 49def1853334 ("Linux 4.10-rc4") 566cf877a1fc ("Linux 4.10-rc6") 5771a8c08880 ("Linux v4.13-rc1") 7089db84e356 ("Linux 4.10-rc8") 7a308bb3016f ("Linux 4.10-rc5") 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 7ce7d89f4883 ("Linux 4.10-rc1") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") a121103c9228 ("Linux 4.10-rc3") a81c42136604 ("perf s390: add regs_query_register_offset()") a9fc2db0a8ab ("s390/perf: define common DWARF register string table") b24413180f56 ("License cleanup: add SPDX GPL-2.0 license identifier to files with no license") c1ae3cfa0e89 ("Linux 4.11-rc1") c470abd4fde4 ("Linux 4.10") d5adbfcd5f7b ("Linux 4.10-rc7")
v4.4.236: Failed to apply! Possible dependencies: 0c4d40d58075 ("tools build: Add BPF feature check to test-all") 1925459b4d92 ("tools build: Fix feature Makefile issues with 'O='") 58683600dfe3 ("perf build: Use FEATURE-DUMP in bpf subproject") 76ee2ff34274 ("tools build feature: Move dwarf post unwind choice output into perf") 7af7919f0f4b ("tools include s390: Grab a copy of arch/s390/include/uapi/asm/unistd.h") 8ee4646038e4 ("perf build: Add libcrypto feature detection") 95f28190aa01 ("tools include arch: Grab a copy of errno.h for arch's supported by perf") 96b9e70b8e6c ("perf build: Introduce FEATURES_DUMP make variable") 9fd4186ac19a ("tools build: Allow subprojects select all feature checkers") abb26210a395 ("perf tools: Force fixdep compilation at the start of the build") aeafd623f866 ("perf tools: Move headers check into bash script") c053a1506fae ("perf build: Select all feature checkers for feature-dump") d4dfdf00d43e ("perf jvmti: Plug compilation into perf build") d58ac0bf8d1e ("perf build: Add clang and llvm compile and linking support") d8ad6a15cc3a ("tools lib bpf: Don't do a feature check when cleaning") e12b202f8fb9 ("perf jitdump: Build only on supported archs") e26e63be64a1 ("perf build: Add sdt feature detection")
NOTE: The patch will not be queued to stable trees until it is upstream.
How should we proceed with this patch?
[cc: linux-ppc, Arnd, Paul]
The patch makes identical changes to 'arch/powerpc/include/uapi/asm/errno.h' and its copy 'tools/arch/powerpc/include/uapi/asm/errno.h' first created in kernel v4.16. Since it's the patch hunk for the latter file which is failing on backports to < v4.16, I would think it OK to skip that hunk where the latter file is missing. I'd prefer to let Michael decide the best course as he's still reviewing the patch.
Thanks, Tony
-- Thanks Sasha
Hello Michael,
The latest version of this patch addressed all feedback I'm aware of when submitted last September, and I've seen no further comments from reviewers since then.
Could you please let me know where this stands and if anything further is needed?
Kind regards, Tony
On Thu, 17 Sept 2020 at 06:54, Tony Ambardar tony.ambardar@gmail.com wrote:
A few archs like powerpc have different errno.h values for macros EDEADLOCK and EDEADLK. In code including both libc and linux versions of errno.h, this can result in multiple definitions of EDEADLOCK in the include chain. Definitions to the same value (e.g. seen with mips) do not raise warnings, but on powerpc there are redefinitions changing the value, which raise warnings and errors (if using "-Werror").
Guard against these redefinitions to avoid build errors like the following, first seen cross-compiling libbpf v5.8.9 for powerpc using GCC 8.4.0 with musl 1.1.24:
In file included from ../../arch/powerpc/include/uapi/asm/errno.h:5, from ../../include/linux/err.h:8, from libbpf.c:29: ../../include/uapi/asm-generic/errno.h:40: error: "EDEADLOCK" redefined [-Werror] #define EDEADLOCK EDEADLK
In file included from toolchain-powerpc_8540_gcc-8.4.0_musl/include/errno.h:10, from libbpf.c:26: toolchain-powerpc_8540_gcc-8.4.0_musl/include/bits/errno.h:58: note: this is the location of the previous definition #define EDEADLOCK 58
cc1: all warnings being treated as errors
CC: Stable stable@vger.kernel.org Reported-by: Rosen Penev rosenp@gmail.com Signed-off-by: Tony Ambardar Tony.Ambardar@gmail.com
v1 -> v2:
- clean up commit description formatting
v2 -> v3: (per Michael Ellerman)
- drop indeterminate 'Fixes' tags, request stable backports instead
arch/powerpc/include/uapi/asm/errno.h | 1 + tools/arch/powerpc/include/uapi/asm/errno.h | 1 + 2 files changed, 2 insertions(+)
diff --git a/arch/powerpc/include/uapi/asm/errno.h b/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/arch/powerpc/include/uapi/asm/errno.h +++ b/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK diff --git a/tools/arch/powerpc/include/uapi/asm/errno.h b/tools/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/tools/arch/powerpc/include/uapi/asm/errno.h +++ b/tools/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK
2.25.1
Tony Ambardar tony.ambardar@gmail.com writes:
Hello Michael,
The latest version of this patch addressed all feedback I'm aware of when submitted last September, and I've seen no further comments from reviewers since then.
Could you please let me know where this stands and if anything further is needed?
Sorry, it's still sitting in my inbox :/
I was going to reply to suggest we split the tools change out. The headers under tools are usually updated by another maintainer, I think it might even be scripted.
Anyway I've applied your patch and done that (dropped the change to tools/.../errno.h), which should also mean the stable backport is more likely to work automatically.
It will hit mainline in v5.13-rc1 and then be backported to the stable trees.
I don't think you actually need the tools version of the header updated to fix your bug? In which case we can probably just wait for it to be updated automatically when the tools headers are sync'ed with the kernel versions.
cheers
On Thu, 17 Sept 2020 at 06:54, Tony Ambardar tony.ambardar@gmail.com wrote:
A few archs like powerpc have different errno.h values for macros EDEADLOCK and EDEADLK. In code including both libc and linux versions of errno.h, this can result in multiple definitions of EDEADLOCK in the include chain. Definitions to the same value (e.g. seen with mips) do not raise warnings, but on powerpc there are redefinitions changing the value, which raise warnings and errors (if using "-Werror").
Guard against these redefinitions to avoid build errors like the following, first seen cross-compiling libbpf v5.8.9 for powerpc using GCC 8.4.0 with musl 1.1.24:
In file included from ../../arch/powerpc/include/uapi/asm/errno.h:5, from ../../include/linux/err.h:8, from libbpf.c:29: ../../include/uapi/asm-generic/errno.h:40: error: "EDEADLOCK" redefined [-Werror] #define EDEADLOCK EDEADLK
In file included from toolchain-powerpc_8540_gcc-8.4.0_musl/include/errno.h:10, from libbpf.c:26: toolchain-powerpc_8540_gcc-8.4.0_musl/include/bits/errno.h:58: note: this is the location of the previous definition #define EDEADLOCK 58
cc1: all warnings being treated as errors
CC: Stable stable@vger.kernel.org Reported-by: Rosen Penev rosenp@gmail.com Signed-off-by: Tony Ambardar Tony.Ambardar@gmail.com
v1 -> v2:
- clean up commit description formatting
v2 -> v3: (per Michael Ellerman)
- drop indeterminate 'Fixes' tags, request stable backports instead
arch/powerpc/include/uapi/asm/errno.h | 1 + tools/arch/powerpc/include/uapi/asm/errno.h | 1 + 2 files changed, 2 insertions(+)
diff --git a/arch/powerpc/include/uapi/asm/errno.h b/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/arch/powerpc/include/uapi/asm/errno.h +++ b/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK diff --git a/tools/arch/powerpc/include/uapi/asm/errno.h b/tools/arch/powerpc/include/uapi/asm/errno.h index cc79856896a1..4ba87de32be0 100644 --- a/tools/arch/powerpc/include/uapi/asm/errno.h +++ b/tools/arch/powerpc/include/uapi/asm/errno.h @@ -2,6 +2,7 @@ #ifndef _ASM_POWERPC_ERRNO_H #define _ASM_POWERPC_ERRNO_H
+#undef EDEADLOCK #include <asm-generic/errno.h>
#undef EDEADLOCK
2.25.1
On Fri, 16 Apr 2021 at 03:41, Michael Ellerman mpe@ellerman.id.au wrote:
Tony Ambardar tony.ambardar@gmail.com writes:
Hello Michael,
The latest version of this patch addressed all feedback I'm aware of when submitted last September, and I've seen no further comments from reviewers since then.
Could you please let me know where this stands and if anything further is needed?
Sorry, it's still sitting in my inbox :/
I was going to reply to suggest we split the tools change out. The headers under tools are usually updated by another maintainer, I think it might even be scripted.
Anyway I've applied your patch and done that (dropped the change to tools/.../errno.h), which should also mean the stable backport is more likely to work automatically.
It will hit mainline in v5.13-rc1 and then be backported to the stable trees.
I don't think you actually need the tools version of the header updated to fix your bug? In which case we can probably just wait for it to be updated automatically when the tools headers are sync'ed with the kernel versions.
cheers
I appreciate the follow up. My original bug was indeed with the tools header but is being patched locally, so waiting for those headers to sync with the kernel versions is fine if it simplifies things overall.
Thanks, Tony
On Thu, 17 Sep 2020 06:54:37 -0700, Tony Ambardar wrote:
A few archs like powerpc have different errno.h values for macros EDEADLOCK and EDEADLK. In code including both libc and linux versions of errno.h, this can result in multiple definitions of EDEADLOCK in the include chain. Definitions to the same value (e.g. seen with mips) do not raise warnings, but on powerpc there are redefinitions changing the value, which raise warnings and errors (if using "-Werror").
[...]
Applied to powerpc/next.
[1/1] powerpc: fix EDEADLOCK redefinition error in uapi/asm/errno.h https://git.kernel.org/powerpc/c/7de21e679e6a789f3729e8402bc440b623a28eae
cheers
linux-stable-mirror@lists.linaro.org
-
Michael Ellerman -
Michael Ellerman -
Sasha Levin -
Tony Ambardar