Re: [PATCH v2 13/23] mm: introduce bpf_out_of_memory() BPF kfunc

[Roman Gushchin <roman.gushchin@linux.dev>]

bot+bpf-ci@kernel.org writes:

> ```
> commit bce57b9b96e05731a14c7e10a364fdd9b733cda2
> Author: Roman Gushchin <roman.gushchin@linux.dev>
>
> mm: introduce bpf_out_of_memory() BPF kfunc
>
> This patch introduces bpf_out_of_memory() BPF kfunc to allow BPF programs
> to declare OOM events and trigger the kernel OOM handling mechanism. The
> function accepts a nullable memcg pointer for system-wide or memcg-scoped
> OOMs.
>
>> diff --git a/include/linux/oom.h b/include/linux/oom.h
>> index 721087952..3cbdcd013 100644
>> --- a/include/linux/oom.h
>> +++ b/include/linux/oom.h
>> @@ -21,6 +21,11 @@ enum oom_constraint {
>>  	CONSTRAINT_MEMCG,
>>  };
>>
>> +enum bpf_oom_flags {
>> +	BPF_OOM_FLAGS_WAIT_ON_OOM_LOCK = 1 << 0,
>> +	BPF_OOM_FLAGS_LAST = 1 << 1,
>> +};
>> +
>>  /*
>>   * Details of the page allocation that triggered the oom killer that are used to
>>   * determine what should be killed.
>> diff --git a/mm/oom_kill.c b/mm/oom_kill.c
>> index 3c86cd755..d7fca4bf5 100644
>> --- a/mm/oom_kill.c
>> +++ b/mm/oom_kill.c
>
> [ ... ]
>
>> +/**
>> + * bpf_out_of_memory - declare Out Of Memory state and invoke OOM killer
>> + * @memcg__nullable: memcg or NULL for system-wide OOMs
>> + * @order: order of page which wasn't allocated
>> + * @flags: flags
>> + * @constraint_text__nullable: custom constraint description for the OOM report
>> + *
>> + * Declares the Out Of Memory state and invokes the OOM killer.
>> + *
>> + * OOM handlers are synchronized using the oom_lock mutex. If wait_on_oom_lock
>> + * is true, the function will wait on it. Otherwise it bails out with -EBUSY
>> + * if oom_lock is contended.
>> + *
>> + * Generally it's advised to pass wait_on_oom_lock=false for global OOMs
>> + * and wait_on_oom_lock=true for memcg-scoped OOMs.
>> + *
>> + * Returns 1 if the forward progress was achieved and some memory was freed.
>> + * Returns a negative value if an error occurred.
>> + */
>> +__bpf_kfunc int bpf_out_of_memory(struct mem_cgroup *memcg__nullable,
>> +				  int order, u64 flags)
>> +{
>> +	struct oom_control oc = {
>> +		.memcg = memcg__nullable,
>> +		.order = order,
>> +	};
>
> Can bpf_out_of_memory() work correctly for system-wide OOMs with
> uninitialized gfp_mask?
>
> Looking at out_of_memory():
>
>     if (!(oc->gfp_mask & __GFP_FS) && !is_memcg_oom(oc))
>         return true;
>
> With gfp_mask=0 and memcg=NULL (system-wide OOM), this evaluates to:
> !(0 & __GFP_FS) && !false = true, causing an early return without
> invoking the OOM killer. The comment in out_of_memory() explains this
> check is about allocation context: "The OOM killer does not compensate
> for IO-less reclaim."
>
> All other callers of out_of_memory() initialize gfp_mask:
> - mm/page_alloc.c: .gfp_mask = gfp_mask
> - mm/memcontrol.c: .gfp_mask = gfp_mask
>
> For memcg OOMs, the code works because is_memcg_oom(oc) is true,
> bypassing the check. But for system-wide OOMs (NULL memcg), won't this
> cause the function to return immediately without killing any process?

This is a good catch! It must be .gfp_mask = GFP_KERNEL.

Fixed.

Thanks!