From: Shuah Khan <skhan@linuxfoundation.org>
To: Ingo Molnar <mingo@kernel.org>
Cc: tglx@linutronix.de, mingo@redhat.com, bp@alien8.de,
dave.hansen@linux.intel.com, x86@kernel.org,
linux-kernel@vger.kernel.org,
Shuah Khan <skhan@linuxfoundation.org>,
hpa@zytor.com
Subject: Re: [PATCH] x86/Kconfig: Fix dependency for X86_DEBUG_FPU
Date: Wed, 9 Apr 2025 15:41:51 -0600 [thread overview]
Message-ID: <e0842dd0-71d3-4de0-a2ee-e83493df890b@linuxfoundation.org> (raw)
In-Reply-To: <Z_bnzs3IPyhVIYaT@gmail.com>
On 4/9/25 15:34, Ingo Molnar wrote:
>
> * Shuah Khan <skhan@linuxfoundation.org> wrote:
>
>> On 4/8/25 00:54, Ingo Molnar wrote:
>>>
>>> * Shuah Khan <skhan@linuxfoundation.org> wrote:
>>>
>>>> Compile fails when X86_DEBUG_FPU is enabled and X86_REQUIRED_FEATURE_FPU is
>>>> disabled. Add explicit dependency on X86_REQUIRED_FEATURE_FPU to fix it.
>>>>
>>>> ../arch/x86/kernel/fpu/regset.c:411:(.text+0x4cf2f): undefined reference to `fpregs_soft_get'
>>>> ld: vmlinux.o: in function `fpregs_set':
>>>> ../arch/x86/kernel/fpu/regset.c:445:(.text+0x4d0da): undefined reference to `fpregs_soft_set'
>>>> ld: vmlinux.o: in function `copy_fpstate_to_sigframe':
>>>> ../arch/x86/kernel/fpu/signal.c:197:(.text+0x4da98): undefined reference to `fpregs_soft_get'
>>>>
>>>
>>> So I cannot reproduce this build failure on either v6.14 or v6.15-rc1:
>>>
>>> starship:~/tip> git describe
>>> v6.15-rc1
>>>
>>> starship:~/tip> grep -E 'X86_32=|M486SX=|X86_REQUIRED_FEATURE|X86_DEBUG_FPU' .config
>>> CONFIG_X86_32=y
>>> CONFIG_M486SX=y
>>> CONFIG_X86_REQUIRED_FEATURE_ALWAYS=y
>>> CONFIG_X86_DEBUG_FPU=y
>>>
>>> starship:~/tip> make -j128 ARCH=i386 bzImage modules
>>>
>>> ...
>>> Kernel: arch/x86/boot/bzImage is ready (#5)
>>>
>>>> Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
>>>> ---
>>>> 6.14 compile worked without X86_REQUIRED_FEATURE_FPU enabled. Might be a
>>>> new dependency. Enabling X86_REQUIRED_FEATURE_FPU fixed it for me on my
>>>> test system.
Right. That is what I meant by new dependency
>>>
>>> So vanilla v6.14 does not have X86_REQUIRED_FEATURE Kconfig flags, at
>>> all:
>>>
>>> starship:~/tip> grep FPU .config
>>> CONFIG_ARCH_HAS_KERNEL_FPU_SUPPORT=y
>>> CONFIG_X86_DEBUG_FPU=y
>>> # CONFIG_TEST_FPU is not set
>>>
>>> Because X86_REQUIRED_FEATURE_FPU is a new v6.15-rc1 feature, introduced
>>> via:
>>>
>>> 3d37d9396eb3 ("x86/cpufeatures: Add {REQUIRED,DISABLED} feature configs")
>>>
>>> I have no doubt you are seeing this build failure - but I think there
>>> might be some other .config detail required to reproduce it, not
>>> mentioned in your changelog. Could you please send the config you used?
>>>
>>
>> Config attached.
>>
>> grep _FPU /boot/config-6.14.0+
>> CONFIG_ARCH_HAS_KERNEL_FPU_SUPPORT=y
>> CONFIG_X86_DEBUG_FPU=y
>> # CONFIG_TEST_FPU is not set
>>
>> I noticed you are building ARCH=i386 - can you reproduce this
>> with x86_64? That is what I am building.
>>
>> thanks,
>> -- Shuah
>
>> #
>> # Automatically generated file; DO NOT EDIT.
>> # Linux/x86 6.14.0 Kernel Configuration
>> #
>> CONFIG_CC_VERSION_TEXT="gcc (Debian 14.2.0-17) 14.2.0"
>
> Is this a vanilla v6.14 kernel? Because on upstream v6.14 'make
> oldconfig' gives me:
>
> Generate BTF type information (DEBUG_INFO_BTF) [N/y/?] (NEW)
>
> Maybe different build environment?
>
> ... and after answering that with 'N' the build succeeds:
>
> LD arch/x86/boot/setup.elf
> OBJCOPY arch/x86/boot/setup.bin
> BUILD arch/x86/boot/bzImage
> Kernel: arch/x86/boot/bzImage is ready (#3)
Yes it is vanilla
>
> More importantly, X86_REQUIRED_FEATURE_FPU *does not exist* in the
> vanilla v6.14 kernel, it's a new v6.15 feature. So this part of your
> changelog totally doesn't apply to a v6.14 kernel:
I started with vanilla 6.14 kernel running oldconfig on it.
In this case if X86_DEBUG_FPU is enabled in the oldconfig,
should the config generated for 6.15 add X86_REQUIRED_FEATURE_FPU.
It appears there is a dependency between X86_DEBUG_FPU and
the newly added X86_REQUIRED_FEATURE_FPU
thanks,
-- Shuah
next prev parent reply other threads:[~2025-04-09 21:41 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-07 23:10 [PATCH] x86/Kconfig: Fix dependency for X86_DEBUG_FPU Shuah Khan
2025-04-08 6:54 ` Ingo Molnar
2025-04-08 16:51 ` Shuah Khan
2025-04-09 21:34 ` Ingo Molnar
2025-04-09 21:41 ` Shuah Khan [this message]
2025-04-10 6:31 ` Ingo Molnar
2025-04-10 21:18 ` Shuah Khan
2025-04-13 19:04 ` Ingo Molnar
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=e0842dd0-71d3-4de0-a2ee-e83493df890b@linuxfoundation.org \
--to=skhan@linuxfoundation.org \
--cc=bp@alien8.de \
--cc=dave.hansen@linux.intel.com \
--cc=hpa@zytor.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/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