❌ FAIL (MISSED 2 of 87): Test report for for-kernelci (6.9.0-rc3, arm-next, fec50db7)

[Catalin Marinas]

On Wed, Apr 10, 2024 at 02:54:39PM +0100, Will Deacon wrote:
> On Tue, Apr 09, 2024 at 06:01:06PM -0000, cki-project at redhat.com wrote:
> > Hi, we tested your kernel and here are the results:
> > 
> >     Overall result: FAILED
> >              Merge: OK
> >            Compile: OK
> >               Test: FAILED
> > 
> > 
> > Kernel information:
> >     Commit message: Linux 6.9-rc3
> > 
> > You can find all the details about the test run at
> >     https://datawarehouse.cki-project.org/kcidb/checkouts/139776
> > 
> > One or more kernel tests failed:
> >     Unrecognized or new issues:
> >         Boot test
> >              aarch64
> >                    Logs: https://datawarehouse.cki-project.org/kcidb/tests/11963290
> >                    Non-passing ran subtests:
> >                        ❌ FAIL distribution/kpkginstall/journalctl-check
> 
> This is just with a vanilla -rc3 kernel to get a baseline result before
> queuing patches for 6.10, but the failure seems to stem from a GFP_DMA32
> allocation failure in the mellanox driver:
> 
>  | (udev-worker): page allocation failure: order:0, mode:0xcc4(GFP_KERNEL|GFP_DMA32), nodemask=(null),cpuset=/,mems_allowed=0
>  | CPU: 24 PID: 927 Comm: (udev-worker) Not tainted 6.9.0-rc3 #1
>  | Hardware name: Lenovo HR330A            7X33CTO1WW    /FALCON     , BIOS hve104r-1.15 02/26/2021
>  | Call trace:
>  |  dump_backtrace+0xdc/0x140
>  |  show_stack+0x20/0x40
>  |  dump_stack_lvl+0x60/0x80
>  |  dump_stack+0x18/0x28
>  |  warn_alloc+0x138/0x1c8
>  |  __alloc_pages_slowpath.constprop.0+0x89c/0x958
>  |  __alloc_pages+0x2cc/0x360
>  |  __dma_direct_alloc_pages.isra.0+0x118/0x290
>  |  dma_direct_alloc+0x21c/0x318
>  |  dma_alloc_attrs+0x90/0x110
>  |  mlx4_create_eq+0x160/0x4c8 [mlx4_core]
[...]

If I read the rest of the log correctly [1], there's plenty of memory in
ZONE_DMA:

| Node 0 DMA free:1035920kB boost:0kB min:452kB low:1752kB high:3052kB
| reserved_highatomic:0KB active_anon:0kB inactive_anon:0kB
| active_file:0kB inactive_file:0kB unevictable:0kB writepending:0kB
| present:1835008kB managed:1301248kB mlocked:0kB bounce:0kB
| free_pcp:3144kB local_pcp:3144kB free_cma:1035920kB 

For some reason the GFP_DMA32 allocation doesn't fall through to
ZONE_DMA.

[1] https://s3.amazonaws.com/arr-cki-prod-trusted-artifacts/trusted-artifacts/1245911266/test_aarch64/6582401556/artifacts/run.done.01/results_0001/console.log

-- 
Catalin