Re: [PATCH 5.15 000/215] 5.15.166-rc1 review

[Greg KH <gregkh@linuxfoundation.org>]

On Mon, Sep 02, 2024 at 03:39:49PM -0700, Richard Narron wrote:
> I get an "out of memory" error when building Linux kernels 5.15.164,
> 5.15.165 and 5.15.166-rc1:
> ...
>   LD [M]  drivers/mtd/tests/mtd_stresstest.o
>   LD [M]  drivers/pcmcia/pcmcia_core.o
>   LD [M]  drivers/mtd/tests/mtd_subpagetest.o
> 
> cc1: out of memory allocating 180705472 bytes after a total of 283914240
> bytes
>   LD [M]  drivers/mtd/tests/mtd_torturetest.o
>   CC [M]  drivers/mtd/ubi/wl.o
>   LD [M]  drivers/pcmcia/pcmcia.o
>   CC [M]  drivers/gpu/drm/nouveau/nvkm/engine/disp/headgv100.o
>   CC [M]  drivers/gpu/drm/amd/amdgpu/../display/dc/dce/dmub_hw_lock_mgr.o
>   LD [M]  drivers/mtd/tests/mtd_nandbiterrs.o
>   CC [M]  drivers/mtd/ubi/attach.o
>   LD [M]  drivers/staging/qlge/qlge.o
> make[4]: *** [scripts/Makefile.build:289:
> drivers/staging/media/atomisp/pci/isp/kernels/ynr/ynr_1.0/ia_css_ynr.host.o]
> Error 1
> make[3]: *** [scripts/Makefile.build:552: drivers/staging/media/atomisp]
> Error 2
> make[2]: *** [scripts/Makefile.build:552: drivers/staging/media] Error 2
> make[2]: *** Waiting for unfinished jobs....
>   LD [M]  drivers/pcmcia/pcmcia_rsrc.o
>   CC [M]  drivers/gpu/drm/amd/amdgpu/../display/dc/dce/dmub_outbox.o
> make[1]: *** [scripts/Makefile.build:552: drivers/staging] Error 2
> make[1]: *** Waiting for unfinished jobs....
>   CC [M]  drivers/gpu/drm/amd/amdgpu/../display/dc/dml/calcs/dce_calcs.o
>   CC [M]  drivers/gpu/drm/amd/amdgpu/../display/dc/dml/calcs/custom_float.o
>   CC [M]  drivers/net/ethernet/mellanox/mlx5/core/ipoib/ipoib.o
> ...
> 
> #uname -a
> Linux aragorn 5.15.166-rc1-smp #1 SMP PREEMPT Mon Sep 2 14:03:00 PDT 2024
> i686 AMD Ryzen 9 5900X 12-Core Processor AuthenticAMD GNU/Linux
> 
> Attached is my config file.
> 
> I found a work around for this problem.
> 
> Remove the six minmax patches introduced with kernel 5.15.164:
> 
> minmax: allow comparisons of 'int' against 'unsigned char/short'
> minmax: allow min()/max()/clamp() if the arguments have the same
> minmax: clamp more efficiently by avoiding extra comparison
> minmax: fix header inclusions
> minmax: relax check to allow comparison between unsigned arguments
> minmax: sanity check constant bounds when clamping
> 
> Can these 6 patches be removed or fixed?

It's a bit late, as we rely on them for other changes.

Perhaps just fixes for the files that you are seeing build crashes on?
I know a bunch of them went into Linus's tree for this issue, but we
didn't backport them as I didn't know what was, and was not, needed.  If
you can pinpoint the files that cause crashes, I can dig them up.

thanks,

greg k-h