Re: drivers/gpu/drm/amd/amdgpu/../display/dc/dml2/display_mode_core.c:6193:12: warning: stack frame size (3368) exceeds limit (2048) in 'dml_core_mode_support'

[Alexander Lobakin]

From: Yujie Liu <yujie.liu@xxxxxxxxx>
Date: Wed, 7 Feb 2024 11:09:36 +0800

> On Tue, Feb 06, 2024 at 04:18:55PM +0000, Pillai, Aurabindo wrote:
>> [AMD Official Use Only - General]
>>
>> Hi,
>>
>> I tried with LLVM16 (from Manjaro) and couldnt repro the issue on both mainline and amd-staging-drm-next. Could someone share the config used for this build ?
> 
> Hi,
> 
> There is a "reproduce" link in the original report. Hope it would be
> helpful:
> 
> reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20231120/202311200955.pmjdQEM1-lkp@xxxxxxxxx/reproduce)
> 
> For a quick check, please kindly try the following steps to reproduce
> the issue:
> 
> $ git checkout v6.8-rc3
> $ wget https://download.01.org/0day-ci/archive/20231120/202311200955.pmjdQEM1-lkp@xxxxxxxxx/config -O .config
> $ make W=1 LLVM=1 ARCH=x86_64 olddefconfig
> $ make -j$(nproc) W=1 LLVM=1 ARCH=x86_64 drivers/gpu/drm/amd/amdgpu/

My repro is:

1. Clone and build [0]. If using Manjaro or anything else Arch-based,
   you can install llvm-git from AUR.
2. Clone the latest net-next (or mainline or linux-next or whatever).
3. Use some default distro kernel .config -- I'm using Fedora's,
   Arch/Manjaro/etc. would also work.
4. make -j$(($(nproc) + 1)) LLVM=/path/to/built/llvm-git/bin/ LLVM_IAS=1
   all

[0] https://github.com/llvm/llvm-project

Thanks,
Olek