llvm.lists.linux.dev archive mirror
 help / color / mirror / Atom feed
From: Nick Desaulniers <ndesaulniers@google.com>
To: Shreeya Patel <shreeya.patel@collabora.com>
Cc: Michal Marek <michal.lkml@markovi.net>,
	Masahiro Yamada <masahiroy@kernel.org>,
	 Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	clang-built-linux <llvm@lists.linux.dev>,
	 Bill Wendling <morbo@google.com>,
	Nathan Chancellor <nathan@kernel.org>,
	regressions@lists.linux.dev,
	 "gustavo.padovan@collabora.com" <gustavo.padovan@collabora.com>,
	 Guillaume Charles Tucker <guillaume.tucker@collabora.com>,
	denys.f@collabora.com,  ricardo.canuelo@collabora.com,
	kernelci@lists.linux.dev
Subject: Re: [PATCH v4] Makefile.compiler: replace cc-ifversion with compiler-specific macros
Date: Wed, 3 May 2023 14:15:07 -0700	[thread overview]
Message-ID: <CAKwvOd=4hBcU4fAkddU0b-GOZc9FzTZoj3PFW6ZZrX0jS8x+bg@mail.gmail.com> (raw)
In-Reply-To: <89961dfc-d40f-78e4-5d34-b86b7d152182@collabora.com>

On Wed, May 3, 2023 at 2:02 PM Shreeya Patel
<shreeya.patel@collabora.com> wrote:
>
> Hi Nick,
>
> On 28/04/23 22:57, Nick Desaulniers wrote:
> > On Thu, Apr 27, 2023 at 4:54 AM Shreeya Patel
> > <shreeya.patel@collabora.com> wrote:
> >> Hi Nick,
> >>
> >> On 19/09/22 22:38, Nick Desaulniers wrote:
> >>> cc-ifversion is GCC specific. Replace it with compiler specific
> >>> variants. Update the users of cc-ifversion to use these new macros.
> >>>
> >>> Link: https://github.com/ClangBuiltLinux/linux/issues/350
> >>> Link: https://lore.kernel.org/llvm/CAGG=3QWSAUakO42kubrCap8fp-gm1ERJJAYXTnP1iHk_wrH=BQ@mail.gmail.com/
> >>> Suggested-by: Bill Wendling <morbo@google.com>
> >>> Reviewed-by: Nathan Chancellor <nathan@kernel.org>
> >>> Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
> >> KernelCI found this patch causes a regression in the
> >> baseline.logintest on qemu_arm-virt-gicv3-uefi [1],
> >> see the bisection report for more details [2].
> >>
> >> Let me know if you have any questions.
> >>
> >>
> >> [1] https://linux.kernelci.org/test/case/id/644596a0beca2ead032e8669/
> > Hi Shreeya,
> > Thanks for the report.
> >
> > When I click the above link, then click `multi_v7_defconfig+debug` to
> > get the config necessary to reproduce, I get an HTTP 404.
> > https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/kernel.config
> >
> > Same for zImage
> > https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/zImage
>
> Apologies for the broken links. We will try to fix the important ones if
> we can but in the meantime,
> following is the correct link that you can refer.
>
> config :-
> https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/config/kernel.config
>
> zImage :-
> https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/kernel/zImage
>
> If you notice, they are present under the kernel directory and same way
> you can find links for other kernel
> builds if you'd like to check them out.
>
> > If I click on the log
> > https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/lab-collabora/baseline-qemu_arm-virt-gicv3-uefi.txt
> > It looks like the machine powered up, then powered off. Is the test
> > actually failing?
>
> I recommend checking the html logs from the kernelci dashboard.
> Also, FYI baseline.login test failure means that the device failed to
> boot which I think is causing by the issues that you pointed out.
>
> <3>[    0.417001][    T1] UBSAN: array-index-out-of-bounds in
> ../arch/arm/mach-sunxi/mc_smp.c:811:29
>
> And potentially another issue with ftrace
>
> <4>[    0.000000][    T0] WARNING: CPU: 0 PID: 0 at
> kernel/trace/ftrace.c:2176 ftrace_bug+0x340/0x3b4
>
>
> Let me know if you need more information from my side to reproduce this
> on your end.
>
>
> Thanks,
> Shreeya Patel

Hi Shreeya,
I may need your help to reproduce the failure.

$ wget https://storage.kernelci.org/mainline/master/v6.3/arm/multi_v7_defconfig+debug/gcc-10/config/kernel.config
-O .config
$ ARCH=arm CROSS_COMPILE=arm-linux-gnueabihf- make -j128 olddefconfig all -s
<launch qemu>
/ # mount -t proc /proc
/ # cat /proc/version
Linux version 6.3.0 (root@61385772abae) (arm-linux-gnueabihf-gcc
(Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian)
2.35.2) #2 SMP Wed May  3 21:10:27 UTC 2023

I was able to boot the resulting kernel to a command line.  Perhaps
there's something about the userspace image that tickles this? Can you
supply the rootfs that's used in testing?

> > I was able to boot ARCH=arm defconfig with CC=arm-linux-gnueabihf-gcc
> > (Debian 10.2.1-6) in QEMU just fine.  So I'm going to need some more
> > information to help reproduce what specifically is failing.
> >
> > Linux version 6.3.0 (root@61385772abae) (arm-linux-gnueabihf-gcc
> > (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU Binutils for Debian)
> > 2.35.2) #1 SMP Fri Apr 28 17:19:59 UTC 2023
> >
> > ---
> >
> > It does look like UBSAN is flagging an array OOB:
> >
> > <3>[    0.417001][    T1] UBSAN: array-index-out-of-bounds in
> > ../arch/arm/mach-sunxi/mc_smp.c:811:29
> >
> > And potentially another issue with ftrace
> >
> > <4>[    0.000000][    T0] WARNING: CPU: 0 PID: 0 at
> > kernel/trace/ftrace.c:2176 ftrace_bug+0x340/0x3b4
> >
> >
> >
> >> [2] https://groups.io/g/kernelci-results/message/40804
> >>
> >>
> >> Thanks,
> >> Shreeya Patel
> >>
> >> #regzbot introduced: 88b61e3bff93
> >>
> >>> ---
> >>> Changes v3 -> v4:
> >>> * Split into its own patch again from series, as per Masahiro.
> >>> * Rebase on top of b0839b281c427e844143dba3893e25c83cdd6c17 and update
> >>>     clang -Wformat logic in scripts/Makefile.extrawarn, as per Masahiro.
> >>>
> >>>    Documentation/kbuild/makefiles.rst          | 29 ++++++++++++---------
> >>>    Makefile                                    |  6 ++---
> >>>    drivers/gpu/drm/amd/display/dc/dml/Makefile |  2 +-
> >>>    scripts/Makefile.compiler                   | 10 ++++---
> >>>    scripts/Makefile.extrawarn                  |  4 +--
> >>>    5 files changed, 29 insertions(+), 22 deletions(-)
> >>>
> >>> diff --git a/Documentation/kbuild/makefiles.rst b/Documentation/kbuild/makefiles.rst
> >>> index 11a296e52d68..ee7e3ea1fbe1 100644
> >>> --- a/Documentation/kbuild/makefiles.rst
> >>> +++ b/Documentation/kbuild/makefiles.rst
> >>> @@ -682,22 +682,27 @@ more details, with real examples.
> >>>        In the above example, -Wno-unused-but-set-variable will be added to
> >>>        KBUILD_CFLAGS only if gcc really accepts it.
> >>>
> >>> -    cc-ifversion
> >>> -     cc-ifversion tests the version of $(CC) and equals the fourth parameter
> >>> -     if version expression is true, or the fifth (if given) if the version
> >>> -     expression is false.
> >>> +    gcc-min-version
> >>> +     gcc-min-version tests if the value of $(CONFIG_GCC_VERSION) is greater than
> >>> +     or equal to the provided value and evaluates to y if so.
> >>>
> >>>        Example::
> >>>
> >>> -             #fs/reiserfs/Makefile
> >>> -             ccflags-y := $(call cc-ifversion, -lt, 0402, -O1)
> >>> +             cflags-$(call gcc-min-version, 70100) := -foo
> >>>
> >>> -     In this example, ccflags-y will be assigned the value -O1 if the
> >>> -     $(CC) version is less than 4.2.
> >>> -     cc-ifversion takes all the shell operators:
> >>> -     -eq, -ne, -lt, -le, -gt, and -ge
> >>> -     The third parameter may be a text as in this example, but it may also
> >>> -     be an expanded variable or a macro.
> >>> +     In this example, cflags-y will be assigned the value -foo if $(CC) is gcc and
> >>> +     $(CONFIG_GCC_VERSION) is >= 7.1.
> >>> +
> >>> +    clang-min-version
> >>> +     clang-min-version tests if the value of $(CONFIG_CLANG_VERSION) is greater
> >>> +     than or equal to the provided value and evaluates to y if so.
> >>> +
> >>> +     Example::
> >>> +
> >>> +             cflags-$(call clang-min-version, 110000) := -foo
> >>> +
> >>> +     In this example, cflags-y will be assigned the value -foo if $(CC) is clang
> >>> +     and $(CONFIG_CLANG_VERSION) is >= 11.0.0.
> >>>
> >>>        cc-cross-prefix
> >>>        cc-cross-prefix is used to check if there exists a $(CC) in path with
> >>> diff --git a/Makefile b/Makefile
> >>> index 298f69060f10..411c8480b37e 100644
> >>> --- a/Makefile
> >>> +++ b/Makefile
> >>> @@ -790,7 +790,6 @@ KBUILD_CFLAGS += $(stackp-flags-y)
> >>>
> >>>    KBUILD_CFLAGS-$(CONFIG_WERROR) += -Werror
> >>>    KBUILD_CFLAGS-$(CONFIG_CC_NO_ARRAY_BOUNDS) += -Wno-array-bounds
> >>> -KBUILD_CFLAGS += $(KBUILD_CFLAGS-y) $(CONFIG_CC_IMPLICIT_FALLTHROUGH)
> >>>
> >>>    ifdef CONFIG_CC_IS_CLANG
> >>>    KBUILD_CPPFLAGS += -Qunused-arguments
> >>> @@ -972,7 +971,6 @@ ifdef CONFIG_CC_IS_GCC
> >>>    KBUILD_CFLAGS += -Wno-maybe-uninitialized
> >>>    endif
> >>>
> >>> -ifdef CONFIG_CC_IS_GCC
> >>>    # The allocators already balk at large sizes, so silence the compiler
> >>>    # warnings for bounds checks involving those possible values. While
> >>>    # -Wno-alloc-size-larger-than would normally be used here, earlier versions
> >>> @@ -984,8 +982,8 @@ ifdef CONFIG_CC_IS_GCC
> >>>    # ignored, continuing to default to PTRDIFF_MAX. So, left with no other
> >>>    # choice, we must perform a versioned check to disable this warning.
> >>>    # https://lore.kernel.org/lkml/20210824115859.187f272f@canb.auug.org.au
> >>> -KBUILD_CFLAGS += $(call cc-ifversion, -ge, 0901, -Wno-alloc-size-larger-than)
> >>> -endif
> >>> +KBUILD_CFLAGS-$(call gcc-min-version, 90100) += -Wno-alloc-size-larger-than
> >>> +KBUILD_CFLAGS += $(KBUILD_CFLAGS-y) $(CONFIG_CC_IMPLICIT_FALLTHROUGH)
> >>>
> >>>    # disable invalid "can't wrap" optimizations for signed / pointers
> >>>    KBUILD_CFLAGS       += -fno-strict-overflow
> >>> diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
> >>> index cb81ed2fbd53..d70838edba80 100644
> >>> --- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
> >>> +++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
> >>> @@ -34,7 +34,7 @@ dml_ccflags := -mhard-float -maltivec
> >>>    endif
> >>>
> >>>    ifdef CONFIG_CC_IS_GCC
> >>> -ifeq ($(call cc-ifversion, -lt, 0701, y), y)
> >>> +ifneq ($(call gcc-min-version, 70100),y)
> >>>    IS_OLD_GCC = 1
> >>>    endif
> >>>    endif
> >>> diff --git a/scripts/Makefile.compiler b/scripts/Makefile.compiler
> >>> index 94d0d40cddb3..9d18fb91890e 100644
> >>> --- a/scripts/Makefile.compiler
> >>> +++ b/scripts/Makefile.compiler
> >>> @@ -61,9 +61,13 @@ cc-option-yn = $(call try-run,\
> >>>    cc-disable-warning = $(call try-run,\
> >>>        $(CC) -Werror $(KBUILD_CPPFLAGS) $(KBUILD_CFLAGS) -W$(strip $(1)) -c -x c /dev/null -o "$$TMP",-Wno-$(strip $(1)))
> >>>
> >>> -# cc-ifversion
> >>> -# Usage:  EXTRA_CFLAGS += $(call cc-ifversion, -lt, 0402, -O1)
> >>> -cc-ifversion = $(shell [ $(CONFIG_GCC_VERSION)0 $(1) $(2)000 ] && echo $(3) || echo $(4))
> >>> +# gcc-min-version
> >>> +# Usage: cflags-$(call gcc-min-version, 70100) += -foo
> >>> +gcc-min-version = $(shell [ $(CONFIG_GCC_VERSION) -ge $(1) ] && echo y)
> >>> +
> >>> +# clang-min-version
> >>> +# Usage: cflags-$(call clang-min-version, 110000) += -foo
> >>> +clang-min-version = $(shell [ $(CONFIG_CLANG_VERSION) -ge $(1) ] && echo y)
> >>>
> >>>    # ld-option
> >>>    # Usage: KBUILD_LDFLAGS += $(call ld-option, -X, -Y)
> >>> diff --git a/scripts/Makefile.extrawarn b/scripts/Makefile.extrawarn
> >>> index 6ae482158bc4..5769c1939d40 100644
> >>> --- a/scripts/Makefile.extrawarn
> >>> +++ b/scripts/Makefile.extrawarn
> >>> @@ -48,7 +48,7 @@ else
> >>>    ifdef CONFIG_CC_IS_CLANG
> >>>    KBUILD_CFLAGS += -Wno-initializer-overrides
> >>>    # Clang before clang-16 would warn on default argument promotions.
> >>> -ifeq ($(shell [ $(CONFIG_CLANG_VERSION) -lt 160000 ] && echo y),y)
> >>> +ifneq ($(call clang-min-version, 160000),y)
> >>>    # Disable -Wformat
> >>>    KBUILD_CFLAGS += -Wno-format
> >>>    # Then re-enable flags that were part of the -Wformat group that aren't
> >>> @@ -56,7 +56,7 @@ KBUILD_CFLAGS += -Wno-format
> >>>    KBUILD_CFLAGS += -Wformat-extra-args -Wformat-invalid-specifier
> >>>    KBUILD_CFLAGS += -Wformat-zero-length -Wnonnull
> >>>    # Requires clang-12+.
> >>> -ifeq ($(shell [ $(CONFIG_CLANG_VERSION) -ge 120000 ] && echo y),y)
> >>> +ifeq ($(call clang-min-version, 120000),y)
> >>>    KBUILD_CFLAGS += -Wformat-insufficient-args
> >>>    endif
> >>>    endif
> >
> >



-- 
Thanks,
~Nick Desaulniers

  reply	other threads:[~2023-05-03 21:15 UTC|newest]

Thread overview: 53+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-07  4:59 [PATCH v3 0/5] fix debug info for asm and DEBUG_INFO_SPLIT Nick Desaulniers
2022-09-07  4:59 ` [PATCH v3 1/5] x86/boot/compressed: prefer cc-option for CFLAGS additions Nick Desaulniers
2022-09-09 21:05   ` Masahiro Yamada
2022-09-07  4:59 ` [PATCH v3 2/5] Makefile.compiler: Use KBUILD_AFLAGS for as-option Nick Desaulniers
2022-09-09 21:06   ` Masahiro Yamada
2022-09-07  4:59 ` [PATCH v3 3/5] Makefile.compiler: replace cc-ifversion with compiler-specific macros Nick Desaulniers
2022-09-09 21:59   ` Masahiro Yamada
2022-09-07  4:59 ` [PATCH v3 4/5] Makefile.debug: re-enable debug info for .S files Nick Desaulniers
2022-09-07  4:59 ` [PATCH v3 5/5] Makefile.debug: set -g unconditional on CONFIG_DEBUG_INFO_SPLIT Nick Desaulniers
2022-09-09 20:56   ` Masahiro Yamada
2022-09-09 21:02     ` Masahiro Yamada
2022-09-19 17:08       ` [PATCH v4] Makefile.compiler: replace cc-ifversion with compiler-specific macros Nick Desaulniers
2022-09-23 19:44         ` Masahiro Yamada
2022-09-24 14:28           ` Masahiro Yamada
2022-09-25  1:22             ` Masahiro Yamada
2022-09-26 21:41               ` Nick Desaulniers
2023-04-27 11:53         ` Shreeya Patel
2023-04-28  7:41           ` Thorsten Leemhuis
2023-05-02  9:48             ` Shreeya Patel
2023-05-02 11:58               ` Linux regression tracking (Thorsten Leemhuis)
2023-04-28 17:27           ` Nick Desaulniers
2023-05-03 21:02             ` Shreeya Patel
2023-05-03 21:15               ` Nick Desaulniers [this message]
2023-05-03 22:33                 ` Shreeya Patel
2023-05-15 23:01                   ` Nick Desaulniers
2023-05-17  8:34                     ` Shreeya Patel
2023-05-17 15:39                     ` Ricardo Cañuelo
2023-05-17 16:27                       ` Nick Desaulniers
2023-05-18 14:23                         ` Ricardo Cañuelo
2023-05-18 21:12                           ` Nick Desaulniers
2023-05-19  8:35                             ` Ricardo Cañuelo
2023-05-19 15:57                               ` Nick Desaulniers
2023-05-22 10:09                                 ` Ricardo Cañuelo
2023-05-22 16:52                                   ` Greg KH
2023-05-22 19:52                                     ` Nick Desaulniers
2023-05-22 20:01                                       ` Greg KH
2023-05-22 20:16                                         ` Nick Desaulniers
2023-05-23 10:27                                       ` Shreeya Patel
2023-05-23 21:27                                         ` Nick Desaulniers
2023-06-12 10:10                                           ` Shreeya Patel
2023-06-20  4:19                                             ` Masahiro Yamada
2023-07-10 12:09                                               ` Linux regression tracking (Thorsten Leemhuis)
2023-07-11 11:16                                                 ` Shreeya Patel
2023-08-29 11:28                                                   ` Linux regression tracking (Thorsten Leemhuis)
2023-09-11 10:05                                                     ` Thorsten Leemhuis
2023-09-15  9:33                                                       ` Shreeya Patel
2023-09-30 10:24                                                         ` Masahiro Yamada
2022-09-19 17:30       ` [PATCH v4] Makefile.debug: set -g unconditional on CONFIG_DEBUG_INFO_SPLIT Nick Desaulniers
2022-09-24  1:34         ` Masahiro Yamada
2022-09-19 17:45       ` [PATCH v4] Makefile.debug: re-enable debug info for .S files Nick Desaulniers
2022-09-24  2:11         ` Masahiro Yamada
2022-09-24  2:20           ` Nick Desaulniers
2022-09-24  6:43             ` Masahiro Yamada

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='CAKwvOd=4hBcU4fAkddU0b-GOZc9FzTZoj3PFW6ZZrX0jS8x+bg@mail.gmail.com' \
    --to=ndesaulniers@google.com \
    --cc=denys.f@collabora.com \
    --cc=guillaume.tucker@collabora.com \
    --cc=gustavo.padovan@collabora.com \
    --cc=kernelci@lists.linux.dev \
    --cc=linux-kernel@vger.kernel.org \
    --cc=llvm@lists.linux.dev \
    --cc=masahiroy@kernel.org \
    --cc=michal.lkml@markovi.net \
    --cc=morbo@google.com \
    --cc=nathan@kernel.org \
    --cc=regressions@lists.linux.dev \
    --cc=ricardo.canuelo@collabora.com \
    --cc=shreeya.patel@collabora.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).