Re: [PATCH] recordmcount: Support empty section from recent binutils
Le 29/03/2022 à 00:31, Joel Stanley a écrit : > On Mon, 29 Nov 2021 at 22:43, Christophe Leroy > wrote: >> >> >> >> Le 29/11/2021 à 18:43, Steven Rostedt a écrit : >>> On Fri, 26 Nov 2021 08:43:23 + >>> LEROY Christophe wrote: >>> Le 24/11/2021 à 15:43, Christophe Leroy a écrit : > Looks like recent binutils (2.36 and over ?) may empty some section, > leading to failure like: > > Cannot find symbol for section 11: .text.unlikely. > kernel/kexec_file.o: failed > make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error > 1 > > In order to avoid that, ensure that the section has a content before > returning it's name in has_rel_mcount(). This patch doesn't work, on PPC32 I get the following message with this patch applied: [0.00] ftrace: No functions to be traced? Without the patch I get: [0.00] ftrace: allocating 22381 entries in 66 pages [0.00] ftrace: allocated 66 pages with 2 groups >>> >>> Because of this report, I have not applied this patch (even though I was >>> about to push it to Linus). >>> >>> I'm pulling it from my queue until this gets resolved. >>> >> >> I have no idea on how to fix that for the moment. >> >> With GCC 10 (binutils 2.36) an objdump -x on kernel/kexec_file.o gives: >> >> ld .text.unlikely .text.unlikely >> wF .text.unlikely 0038 >> .arch_kexec_apply_relocations_add >> 0038 wF .text.unlikely 0038 >> .arch_kexec_apply_relocations >> >> >> With GCC 11 (binutils 2.37) the same gives: >> >> wF .text.unlikely 0038 >> .arch_kexec_apply_relocations_add >> 0038 wF .text.unlikely 0038 >> .arch_kexec_apply_relocations >> >> >> The problem is that recordmcount drops weak symbols, and it doesn't find >> any non-weak symbol in .text.unlikely >> >> Explication given at >> https://elixir.bootlin.com/linux/v5.16-rc2/source/scripts/recordmcount.h#L506 >> >> I have no idea on what to do. > > Did you end up finding a solution for this issue? > Not really, my solution was to switch to the kernel compiler at https://mirrors.edge.kernel.org/pub/tools/crosstool/ which embeds binutils 2.36 But it looks like using objtool instead of recordmcount doesn't exhibit the problem. https://patchwork.ozlabs.org/project/linuxppc-dev/patch/20220318105140.43914-4...@linux.ibm.com/ Christophe
Re: [PATCH] recordmcount: Support empty section from recent binutils
On Mon, 29 Nov 2021 at 22:43, Christophe Leroy wrote: > > > > Le 29/11/2021 à 18:43, Steven Rostedt a écrit : > > On Fri, 26 Nov 2021 08:43:23 + > > LEROY Christophe wrote: > > > >> Le 24/11/2021 à 15:43, Christophe Leroy a écrit : > >>> Looks like recent binutils (2.36 and over ?) may empty some section, > >>> leading to failure like: > >>> > >>> Cannot find symbol for section 11: .text.unlikely. > >>> kernel/kexec_file.o: failed > >>> make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error 1 > >>> > >>> In order to avoid that, ensure that the section has a content before > >>> returning it's name in has_rel_mcount(). > >> > >> This patch doesn't work, on PPC32 I get the following message with this > >> patch applied: > >> > >> [0.00] ftrace: No functions to be traced? > >> > >> Without the patch I get: > >> > >> [0.00] ftrace: allocating 22381 entries in 66 pages > >> [0.00] ftrace: allocated 66 pages with 2 groups > > > > Because of this report, I have not applied this patch (even though I was > > about to push it to Linus). > > > > I'm pulling it from my queue until this gets resolved. > > > > I have no idea on how to fix that for the moment. > > With GCC 10 (binutils 2.36) an objdump -x on kernel/kexec_file.o gives: > > ld .text.unlikely .text.unlikely > wF .text.unlikely 0038 > .arch_kexec_apply_relocations_add > 0038 wF .text.unlikely 0038 > .arch_kexec_apply_relocations > > > With GCC 11 (binutils 2.37) the same gives: > > wF .text.unlikely 0038 > .arch_kexec_apply_relocations_add > 0038 wF .text.unlikely 0038 > .arch_kexec_apply_relocations > > > The problem is that recordmcount drops weak symbols, and it doesn't find > any non-weak symbol in .text.unlikely > > Explication given at > https://elixir.bootlin.com/linux/v5.16-rc2/source/scripts/recordmcount.h#L506 > > I have no idea on what to do. Did you end up finding a solution for this issue? Cheers, Joel
Re: [PATCH] recordmcount: Support empty section from recent binutils
Le 29/11/2021 à 18:43, Steven Rostedt a écrit : On Fri, 26 Nov 2021 08:43:23 + LEROY Christophe wrote: Le 24/11/2021 à 15:43, Christophe Leroy a écrit : Looks like recent binutils (2.36 and over ?) may empty some section, leading to failure like: Cannot find symbol for section 11: .text.unlikely. kernel/kexec_file.o: failed make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error 1 In order to avoid that, ensure that the section has a content before returning it's name in has_rel_mcount(). This patch doesn't work, on PPC32 I get the following message with this patch applied: [0.00] ftrace: No functions to be traced? Without the patch I get: [0.00] ftrace: allocating 22381 entries in 66 pages [0.00] ftrace: allocated 66 pages with 2 groups Because of this report, I have not applied this patch (even though I was about to push it to Linus). I'm pulling it from my queue until this gets resolved. I have no idea on how to fix that for the moment. With GCC 10 (binutils 2.36) an objdump -x on kernel/kexec_file.o gives: ld .text.unlikely .text.unlikely wF .text.unlikely 0038 .arch_kexec_apply_relocations_add 0038 wF .text.unlikely 0038 .arch_kexec_apply_relocations With GCC 11 (binutils 2.37) the same gives: wF .text.unlikely 0038 .arch_kexec_apply_relocations_add 0038 wF .text.unlikely 0038 .arch_kexec_apply_relocations The problem is that recordmcount drops weak symbols, and it doesn't find any non-weak symbol in .text.unlikely Explication given at https://elixir.bootlin.com/linux/v5.16-rc2/source/scripts/recordmcount.h#L506 I have no idea on what to do. Thanks Christophe
Re: [PATCH] recordmcount: Support empty section from recent binutils
On Fri, 26 Nov 2021 08:43:23 + LEROY Christophe wrote: > Le 24/11/2021 à 15:43, Christophe Leroy a écrit : > > Looks like recent binutils (2.36 and over ?) may empty some section, > > leading to failure like: > > > > Cannot find symbol for section 11: .text.unlikely. > > kernel/kexec_file.o: failed > > make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error 1 > > > > In order to avoid that, ensure that the section has a content before > > returning it's name in has_rel_mcount(). > > This patch doesn't work, on PPC32 I get the following message with this > patch applied: > > [0.00] ftrace: No functions to be traced? > > Without the patch I get: > > [0.00] ftrace: allocating 22381 entries in 66 pages > [0.00] ftrace: allocated 66 pages with 2 groups Because of this report, I have not applied this patch (even though I was about to push it to Linus). I'm pulling it from my queue until this gets resolved. Thanks, -- Steve
Re: [PATCH] recordmcount: Support empty section from recent binutils
Le 24/11/2021 à 15:43, Christophe Leroy a écrit :
> Looks like recent binutils (2.36 and over ?) may empty some section,
> leading to failure like:
>
> Cannot find symbol for section 11: .text.unlikely.
> kernel/kexec_file.o: failed
> make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error 1
>
> In order to avoid that, ensure that the section has a content before
> returning it's name in has_rel_mcount().
This patch doesn't work, on PPC32 I get the following message with this
patch applied:
[0.00] ftrace: No functions to be traced?
Without the patch I get:
[0.00] ftrace: allocating 22381 entries in 66 pages
[0.00] ftrace: allocated 66 pages with 2 groups
Christophe
>
> Suggested-by: Steven Rostedt
> Link: https://github.com/linuxppc/issues/issues/388
> Link: https://lore.kernel.org/all/20210215162209.5e2a4...@gandalf.local.home/
> Signed-off-by: Christophe Leroy
> ---
> scripts/recordmcount.h | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/scripts/recordmcount.h b/scripts/recordmcount.h
> index 1e9baa5c4fc6..cc6600b729ae 100644
> --- a/scripts/recordmcount.h
> +++ b/scripts/recordmcount.h
> @@ -575,6 +575,8 @@ static char const *has_rel_mcount(Elf_Shdr const *const
> relhdr,
> char const *const shstrtab,
> char const *const fname)
> {
> + if (!shdr0->sh_size)
> + return NULL;
> if (w(relhdr->sh_type) != SHT_REL && w(relhdr->sh_type) != SHT_RELA)
> return NULL;
> return __has_rel_mcount(relhdr, shdr0, shstrtab, fname);
>
[PATCH] recordmcount: Support empty section from recent binutils
Looks like recent binutils (2.36 and over ?) may empty some section,
leading to failure like:
Cannot find symbol for section 11: .text.unlikely.
kernel/kexec_file.o: failed
make[1]: *** [scripts/Makefile.build:287: kernel/kexec_file.o] Error 1
In order to avoid that, ensure that the section has a content before
returning it's name in has_rel_mcount().
Suggested-by: Steven Rostedt
Link: https://github.com/linuxppc/issues/issues/388
Link: https://lore.kernel.org/all/20210215162209.5e2a4...@gandalf.local.home/
Signed-off-by: Christophe Leroy
---
scripts/recordmcount.h | 2 ++
1 file changed, 2 insertions(+)
diff --git a/scripts/recordmcount.h b/scripts/recordmcount.h
index 1e9baa5c4fc6..cc6600b729ae 100644
--- a/scripts/recordmcount.h
+++ b/scripts/recordmcount.h
@@ -575,6 +575,8 @@ static char const *has_rel_mcount(Elf_Shdr const *const
relhdr,
char const *const shstrtab,
char const *const fname)
{
+ if (!shdr0->sh_size)
+ return NULL;
if (w(relhdr->sh_type) != SHT_REL && w(relhdr->sh_type) != SHT_RELA)
return NULL;
return __has_rel_mcount(relhdr, shdr0, shstrtab, fname);
--
2.33.1
