Re: [RFC] Don't create sframes during build

[Indu Bhagat <indu.bhagat@oracle.com>]

On 1/24/26 11:08 AM, Indu wrote:
> On 2026-01-24 4:04 a.m., Sven Joachim wrote:
>> On 2025-09-04 10:14 -0700, Josh Poimboeuf wrote:
>>
>>> On Thu, Sep 04, 2025 at 06:34:04PM +0200, Sebastian Andrzej Siewior 
>>> wrote:
>>>> On 2025-09-04 16:02:42 [+0200], Matthias Klose wrote:
>>>>> [ CCing binutils@sourceware.org ]
>>>>>
>>>>> On 9/4/25 15:18, Sebastian Andrzej Siewior wrote:
>>>>>> Hi,
>>>>>>
>>>>>> gcc in Debian, starting with 15.2.0-2, 14.3.0-6 enables sframe
>>>>>> generation. Unless options like -ffreestanding are passed. Since this
>>>>>> isn't done, there are a few warnings during compile
>>>>>
>>>>> If there are other options when sframe shouldn't be enabled, please 
>>>>> tell.
>>>>
>>>> No, I think this is okay.
>>>>
>>>> …
>>>>>> We could drop the sframe during the final link but this does not 
>>>>>> get rid
>>>>>> of the objtool warnings so we would have to ignore them. But we don't
>>>>>> need it. So what about the following:
>>>>>>
>>>>>> diff --git a/Makefile b/Makefile
>>>>>> --- a/Makefile
>>>>>> +++ b/Makefile
>>>>>> @@ -886,6 +886,8 @@ ifdef CONFIG_CC_IS_GCC
>>>>>>    KBUILD_CFLAGS    += $(call cc-option,--param=allow-store-data- 
>>>>>> races=0)
>>>>>>    KBUILD_CFLAGS    += $(call cc-option,-fno-allow-store-data-races)
>>>>>>    endif
>>>>>> +# No sframe generation for kernel if enabled by default
>>>>>> +KBUILD_CFLAGS    += $(call cc-option,-Xassembler --gsframe=no)
>>>>>>    ifdef CONFIG_READABLE_ASM
>>>>>>    # Disable optimizations that make assembler listings hard to read.
>>>>> This is what I chose for package builds that need disablement of 
>>>>> sframe.
>>>>
>>>> I think this would work for now. Longterm we would have to allow sframe
>>>> creation and keep section if an architecture decides to use it for its
>>>> backtracing. While orc seems fine on x86, there are arm64 patches to 
>>>> use
>>>> for as a stack unwinder.
>>>
>>> This is probably fine, but... how does this interact with other kernel
>>> makefiles enabling sframe?  For example, x86 will soon have a patch to
>>> enable sframe generation for vdso.  And as you mentioned, arm64 will
>>> enable it kernel-wide.
>>>
>>> Removing the objtool !ENDBR warnings would be trivial (and is a good
>>> thing to do regardless).
>>
>> What is the status of sframe support in the kernel?  With current
>> binutils (version 2.45.50.20260119-1) from Debian I could not build
>> Linux 6.18.7:
>>
>> ,----
>> | # LD      arch/x86/boot/compressed/vmlinux
>> |   ld -m elf_x86_64 --no-ld-generated-unwind-info  -pie  --no- 
>> dynamic-linker --orphan-handling=error -z noexecstack --no-warn-rwx- 
>> segments -T arch/x86/boot/compressed/vmlinux.lds arch/x86/boot/ 
>> compressed/kernel_info.o arch/x86/boot/compressed/head_64.o arch/x86/ 
>> boot/compressed/misc.o arch/x86/boot/compressed/string.o arch/x86/ 
>> boot/compressed/cmdline.o arch/x86/boot/compressed/error.o arch/x86/ 
>> boot/compressed/piggy.o arch/x86/boot/compressed/cpuflags.o arch/x86/ 
>> boot/compressed/early_serial_console.o arch/x86/boot/compressed/ 
>> kaslr.o arch/x86/boot/compressed/ident_map_64.o arch/x86/boot/ 
>> compressed/idt_64.o arch/x86/boot/compressed/idt_handlers_64.o arch/ 
>> x86/boot/compressed/pgtable_64.o arch/x86/boot/compressed/acpi.o arch/ 
>> x86/boot/startup/lib.a -o arch/x86/boot/compressed/vmlinux
>> | /usr/bin/x86_64-linux-gnu-ld.bfd: error: unplaced orphan section 
>> `.sframe' from `arch/x86/boot/compressed/kernel_info.o'
>> | /usr/bin/x86_64-linux-gnu-ld.bfd: error: unplaced orphan section 
>> `.sframe' from `arch/x86/boot/compressed/kernel_info.o'
>> | make[6]: *** [arch/x86/boot/compressed/Makefile:116: arch/x86/boot/ 
>> compressed/vmlinux] Error 1
>> | make[5]: *** [arch/x86/boot/Makefile:96: arch/x86/boot/compressed/ 
>> vmlinux] Error 2
>> `----
>>
> 
> Sorry for the breakage.
> 
> There is a patch that needs to make it upstream
> https://sourceware.org/pipermail/binutils/2026-January/147664.html
> (Its a part of a series under review..)
> 
> I will work on getting this patch committed.
> 
> For immediate relief, please use the new ld option --discard-sframe 
> meanwhile.
> 

This was being tracked via 
https://www.sourceware.org/bugzilla/show_bug.cgi?id=33830.

Should be fixed now on Binutils master and 2.46 branch.

Thanks

>> Did not test mainline yet, but "git log --grep=sframe master" does not
>> show anything interesting.
>>
>> Cheers,
>>         Sven
> 
>