Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.debian.bugs.dist > #1245696 > unrolled thread

Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT)

Started byMatthias Klose <doko@debian.org>
First post2025-05-12 14:30 +0200
Last post2025-05-28 20:00 +0200
Articles 6 — 2 participants

Back to article view | Back to linux.debian.bugs.dist

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.


Contents

  Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Matthias Klose <doko@debian.org> - 2025-05-12 14:30 +0200
    Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Ingo Saitz <ingo@hannover.ccc.de> - 2025-05-12 18:30 +0200
      Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Matthias Klose <doko@debian.org> - 2025-05-13 10:40 +0200
        Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Ingo Saitz <ingo@hannover.ccc.de> - 2025-05-14 00:00 +0200
          Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Matthias Klose <doko@debian.org> - 2025-05-21 20:10 +0200
            Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT) Ingo Saitz <ingo@hannover.ccc.de> - 2025-05-28 20:00 +0200

#1245696 — Bug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT)

FromMatthias Klose <doko@debian.org>
Date2025-05-12 14:30 +0200
SubjectBug#1104745: Acknowledgement (gcc-15 ICE compiling linux kernel 6.14.5 with CONFIG_RANDSTRUCT)
Message-ID<KLAsa-2ABg-5@gated-at.bofh.it>
Control: tags -1 + moreinfo

from the attached preprocessed source ...

please can you report the issue, where you got this plugin from?

// gcc version 15.1.0 (Debian 15.1.0-3)
//
// *** WARNING *** there are active plugins, do not report this as a bug 
unless you can reproduce it without enabling any plugins.
// Event                            | Plugins
// PLUGIN_FINISH_TYPE               | randomize_layout_plugin
// PLUGIN_FINISH_DECL               | randomize_layout_plugin
// PLUGIN_ATTRIBUTES                | randomize_layout_plugin
// PLUGIN_ALL_IPA_PASSES_START      | randomize_layout_plugin
// arch/x86/kernel/cpu/proc.c:174:14: internal compiler error: in 
comptypes_check_enum_int, at c/c-typeck.cc:1516
//   174 | const struct seq_operations cpuinfo_op = {
//       |              ^~~~~~~~~~~~~~
// 0x1cf8479 internal_error(char const*, ...)
// 	../../src/gcc/diagnostic-global-context.cc:517
// 0x1cf85eb fancy_abort(char const*, int, char const*)
// 	../../src/gcc/diagnostic.cc:1749
// 0x750e68 comptypes_check_enum_int(tree_node*, tree_node*, bool*)
// 	../../src/gcc/c/c-typeck.cc:1516
// 0x750e68 comptypes_check_enum_int(tree_node*, tree_node*, bool*)
// 	../../src/gcc/c/c-typeck.cc:1510
// 0x24161fd diagnose_mismatched_decls
// 	../../src/gcc/c/c-decl.cc:2170
// 0x24148ff duplicate_decls
// 	../../src/gcc/c/c-decl.cc:3169
// 0x1d93668 pushdecl(tree_node*)
// 	../../src/gcc/c/c-decl.cc:3376
// 0x1df85c4 start_decl(c_declarator*, c_declspecs*, bool, tree_node*, 
bool, unsigned long*)
// 	../../src/gcc/c/c-decl.cc:5773
// 0x2409a09 c_parser_declaration_or_fndef
// 	../../src/gcc/c/c-parser.cc:2846
// 0x2408fbc c_parser_external_declaration
// 	../../src/gcc/c/c-parser.cc:2105
// 0x2405314 c_parser_translation_unit
// 	../../src/gcc/c/c-parser.cc:1959
// 0x2405314 c_parse_file()
// 	../../src/gcc/c/c-parser.cc:29790
// 0x240237d c_common_parse_file()
// 	../../src/gcc/c-family/c-opts.cc:1379

[toc] | [next] | [standalone]


#1245723

FromIngo Saitz <ingo@hannover.ccc.de>
Date2025-05-12 18:30 +0200
Message-ID<KLEcp-2CWY-1@gated-at.bofh.it>
In reply to#1245696
On Mon, May 12, 2025 at 02:25:24PM +0200, Matthias Klose wrote:
> Control: tags -1 + moreinfo
> 
> from the attached preprocessed source ...
> 
> please can you report the issue, where you got this plugin from?

The plugin is part of the linux kernel compilation process to randomize
structure layout. This plugin was from vanilla 6.14.5 linux kernel sources,
in scripts/gcc-plugins/randomize_layout_plugin.c . It gets build and used
automatically when CONFIG_RANDSTRUCT_PERFORMANCE (or _FULL) is enabled.

    Ingo
-- 
const_cast<long double>(Λ)

[toc] | [prev] | [next] | [standalone]


#1245791

FromMatthias Klose <doko@debian.org>
Date2025-05-13 10:40 +0200
Message-ID<KLTl7-2MIA-5@gated-at.bofh.it>
In reply to#1245723
Control: reassign -1 src:linux

On 12.05.25 18:21, Ingo Saitz wrote:
> On Mon, May 12, 2025 at 02:25:24PM +0200, Matthias Klose wrote:
>> Control: tags -1 + moreinfo
>>
>> from the attached preprocessed source ...
>>
>> please can you report the issue, where you got this plugin from?
> 
> The plugin is part of the linux kernel compilation process to randomize
> structure layout. This plugin was from vanilla 6.14.5 linux kernel sources,
> in scripts/gcc-plugins/randomize_layout_plugin.c . It gets build and used
> automatically when CONFIG_RANDSTRUCT_PERFORMANCE (or _FULL) is enabled.
> 
>      Ingo

according to GCC upstream developers, this should be forwarded upstream 
for linux.

[toc] | [prev] | [next] | [standalone]


#1246344

FromIngo Saitz <ingo@hannover.ccc.de>
Date2025-05-14 00:00 +0200
Message-ID<KM5Pk-2Vhy-7@gated-at.bofh.it>
In reply to#1245791
On Tue, May 13, 2025 at 10:33:09AM +0200, Matthias Klose wrote:
> according to GCC upstream developers, this should be forwarded upstream for
> linux.

Do you have any reference on this, because the ICE does not occur when
building the kernel with a vanilla gcc 15.1.0 build from git. (see
Message #10)

OTOH I do get the ICE with all Debians gcc-15 versions I tested (15.1.0-1 to
15.1.0-4).

    Ingo
-- 
const_cast<long double>(Λ)

[toc] | [prev] | [next] | [standalone]


#1247339

FromMatthias Klose <doko@debian.org>
Date2025-05-21 20:10 +0200
Message-ID<KOW37-4K7B-9@gated-at.bofh.it>
In reply to#1246344
On 13.05.25 23:49, Ingo Saitz wrote:
> On Tue, May 13, 2025 at 10:33:09AM +0200, Matthias Klose wrote:
>> according to GCC upstream developers, this should be forwarded upstream for
>> linux.
> 
> Do you have any reference on this, because the ICE does not occur when
> building the kernel with a vanilla gcc 15.1.0 build from git. (see
> Message #10)
> 
> OTOH I do get the ICE with all Debians gcc-15 versions I tested (15.1.0-1 to
> 15.1.0-4).

- please check with the current upstream branch.

- please check with the same configure options and build
   targets as found in Debian.

[toc] | [prev] | [next] | [standalone]


#1248189

FromIngo Saitz <ingo@hannover.ccc.de>
Date2025-05-28 20:00 +0200
Message-ID<KRtei-6nbn-13@gated-at.bofh.it>
In reply to#1247339

[Multipart message — attachments visible in raw view] — view raw

You are right, this seems indeed to be a kernel bug. The ICE occurs when
gcc (15.1.0) is compiled with --enable-checking=misc (which is included in
--enable-checking=all).

I also simplified the file that produces the ICE to three statements
(see attachment).

On Wed, May 21, 2025 at 07:56:00PM +0200, Matthias Klose wrote:
> - please check with the current upstream branch.
> 
> - please check with the same configure options and build
>   targets as found in Debian.

Thanks for your suggestions.

    Ingo
-- 
const_cast<long double>(Λ)

[toc] | [prev] | [standalone]


Back to top | Article view | linux.debian.bugs.dist


csiph-web