Re: [RFC PATCH 0/6] dma-buf: Phase out is_signaled() race

[Christian König]

Hi guys,

On 10/29/26 11:28, Philipp Stanner wrote:
> On Thu, 2026-10-29 at 11:23 +0100, Lucas Stach wrote:
>> Hi Philipp, hi Christian,
> 
> Jo Lucas
> 
>>
>> Am Donnerstag, dem 29.10.2026 um 10:55 +0100 schrieb Philipp Stanner:
> 
> […]
> 
>> I think we would all be interested in the actual use-case that's
>> sensitive to this kind of latency.

Well everything which wants things to start as early as possible when the fence is signaled and not wait for the callbacks to run is affected.

If I'm not completely mistaken when we start to wait for the callbacks to finish in is_signaled we break more things than we fix because some people relied on timing here.

> Me too.
> 
> Let me use this opportunity to stress that in the future, in DRM,
> whenever we have [to accept] a racy "optimization", both commit
> messages and code comments must document why it exists and who
> requested it with what proof.

That is a really good idea I think.

> 
> Would save us lots of time when those things are discussed >5 years
> later.
> 
>>
>> However, I wonder if this is even an issue anymore with the lock
>> inlined into the fence structure.
> 
> Almost no one uses inline-lock so far. So unfortunately, we might
> regress old drivers, maybe.

Considering that I still get bug reports even for the performance of radeon when we change things I fear that this is true.

Regards,
Christian.

> 
> But sounds as if you're in favor of phasing the locklessness out :)
> 
> 
> P.