<!DOCTYPE html><html><head> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"> </head> <body> Am 06.11.24 um 09:14 schrieb Boris Brezillon: <blockquote type="cite" cite="mid:20241106091436.48687e86@collabora.com"> <pre class="moz-quote-pre" wrap="">On Tue, 5 Nov 2024 09:56:22 -0800 Chia-I Wu <a class="moz-txt-link-rfc2396E" href="mailto:olvaffe@gmail.com"><olvaffe@gmail.com></a> wrote: </pre> <blockquote type="cite"> <pre class="moz-quote-pre" wrap="">On Mon, Nov 4, 2024 at 11:32 PM Christian König <a class="moz-txt-link-rfc2396E" href="mailto:christian.koenig@amd.com"><christian.koenig@amd.com></a> wrote: </pre> <blockquote type="cite"> <pre class="moz-quote-pre" wrap=""> Am 04.11.24 um 22:32 schrieb Chia-I Wu: On Tue, Oct 22, 2024 at 10:24 AM Chia-I Wu <a class="moz-txt-link-rfc2396E" href="mailto:olvaffe@gmail.com"><olvaffe@gmail.com></a> wrote: On Tue, Oct 22, 2024 at 9:53 AM Christian König <a class="moz-txt-link-rfc2396E" href="mailto:christian.koenig@amd.com"><christian.koenig@amd.com></a> wrote: Am 22.10.24 um 18:18 schrieb Chia-I Wu: Userspace might poll a syncobj with the query ioctl. Call dma_fence_enable_sw_signaling to ensure dma_fence_is_signaled returns true in finite time. Wait a second, just querying the fence status is absolutely not guaranteed to return true in finite time. That is well documented on the dma_fence() object. When you want to poll on signaling from userspace you really need to call poll or the wait IOCTL with a zero timeout. That will also return immediately but should enable signaling while doing that. So just querying the status should absolutely *not* enable signaling. That's an intentional separation. I think it depends on what semantics DRM_IOCTL_SYNCOBJ_QUERY should have. Well that's what I pointed out. The behavior of the QUERY IOCTL is based on the behavior of the dma_fence and the later is documented to do exactly what it currently does. If DRM_IOCTL_SYNCOBJ_QUERY is mainly for vulkan timeline semaphores, it is a bit heavy if userspace has to do a DRM_IOCTL_SYNCOBJ_TIMELINE_WAIT before a query. Maybe you misunderstood me, you *only* have to call DRM_IOCTL_SYNCOBJ_TIMELINE_WAIT and *not* _QUERY. The underlying dma_fence_wait_timeout() function is extra optimized so that zero timeout has only minimal overhead. This overhead is actually lower than _QUERY because that one actually queries the driver for the current status while _WAIT just assumes that the driver will signal the fence when ready from an interrupt. </pre> </blockquote> <pre class="moz-quote-pre" wrap=""> The context here is that vkGetSemaphoreCounterValue calls QUERY to get the timeline value. WAIT does not replace QUERY.</pre> </blockquote> </blockquote> Oh, that is a really good argument. <blockquote type="cite" cite="mid:20241106091436.48687e86@collabora.com"> <blockquote type="cite"> <pre class="moz-quote-pre" wrap="">Taking a step back, in the binary (singled/unsignaled) case, a WAIT with zero timeout can get the up-to-date status. But in the timeline case, there is no direct way to get the up-to-date status if QUERY must strictly be a wrapper for dma_fence_is_signaled. It comes back to what was QUERY designed for and can we change it?</pre> </blockquote> </blockquote> Yeah that is a really good point. If _QUERY implements a different functionality than _WAIT than the usual distinction between polling and interrupt based approach can't be applied here. <blockquote type="cite" cite="mid:20241106091436.48687e86@collabora.com"> <blockquote type="cite"> <blockquote type="cite"> <pre class="moz-quote-pre" wrap=""> I filed a Mesa issue, <a class="moz-txt-link-freetext" href="https://gitlab.freedesktop.org/mesa/mesa/-/issues/12094">https://gitlab.freedesktop.org/mesa/mesa/-/issues/12094</a>, and Faith suggested a kernel-side fix as well. Should we reconsider this? Wait a second, you might have an even bigger misconception here. The difference between waiting and querying is usually intentional! This is done so that for example on mobile devices you don't need to enable device interrupts, but rather query in defined intervals. This is a very common design pattern and while I don't know the wording of the Vulkan timeline extension it's quite likely that this is the intended use case. </pre> </blockquote> <pre class="moz-quote-pre" wrap="">Yeah, there are Vulkan CTS tests that query timeline semaphores repeatedly for progress. Those tests can fail because mesa translates the queries directly to the QUERY ioctl. As things are, enable_signaling is a requirement to query up-to-date status no matter the syncobj is binary or a timeline. </pre> </blockquote> <pre class="moz-quote-pre" wrap=""> I kinda agree with Chia-I here. What's the point of querying a timeline syncobj if what we get in return is an outdated sync point? I get that the overhead of enabling signalling exists, but if we stand on this position, that means the QUERY ioctl is not suitable for vkGetSemaphoreCounterValue() unless we first add a WAIT(sync_point=0,timeout=0) to make sure signalling is enabled on all fences contained by the dma_fence_chain backing the timeline syncobj. And this has to be done for each vkGetSemaphoreCounterValue(), because new sync points don't have signalling enabled by default.</pre> </blockquote> The common driver design I know from other operating systems is that you either poll without enabling interrupts or you enable interrupts and wait for the async operation to complete. That distinction is really important for things like mobile devices where interrupts are rather power costly. <blockquote type="cite" cite="mid:20241106091436.48687e86@collabora.com"> <pre class="moz-quote-pre" wrap="">At the very least, we should add a new DRM_SYNCOBJ_QUERY_FLAGS_ flag (DRM_SYNCOBJ_QUERY_FLAGS_REFRESH_STATE?) to combine the enable_signalling and query operations in a single ioctl. If we go for this approach, that means mesa has to support both cases, and pick the most optimal one if the kernel supports it. </pre> </blockquote> Another problem we have here is that dma_fence_enable_signaling() can mean two different things, maybe the documentation is a bit confusing: 1) Enabling interrupts so that we don't need to call the ops->is_signaled() driver callback. 2) Asking preemption fences to actually signal because memory management wants to do something. So when this ops->is_signaled() callback is implemented it shouldn't be necessary to enable signaling to query the state. To summarize: When you call _QUERY you shouldn't get an outdated sync point. When you do this then there is something wrong with the backend driver. Regards, Christian. </body> </html>