Metal (Apple) GPU back-end for Tracy #793

slomp · 2024-05-17T05:35:11Z

(I still need to update the manual, but I'm putting the code here for review to save some time).

The Metal back-end in Tracy operates differently than other GPU back-ends like Vulkan, Direct3D and OpenGL. Specifically, TracyMetalZone() must be placed around the site where a command encoder is created.

This is because not all hardware supports timestamps at command granularity, and can only provide timestamps around an entire command encoder. This accommodates for all tiers of hardware; in the future, variants of TracyMetalZone() will be added to support the habitual command-level granularity of Tracy GPU back-ends.

Metal also imposes a few restrictions that make the process of requesting and collecting queries more complicated in Tracy:

timestamp query buffers are limited to 4096 queries (32KB, where each query is 8 bytes)
when a timestamp query buffer is created, Metal initializes all timestamps with zeroes, and there's no way to reset them back to zero after timestamps get resolved; the only way to clear the timestamps is by allocating a new timestamp query buffer
if a command encoder records no commands and its corresponding command buffer ends up committed to the command queue, Metal will "optimize-away" the encoder along with any timestamp queries associated with it (the timestamp will remain as zero and will never get resolved)

Because of the limitations above, two timestamp buffers are managed internally. Once one of the buffers fills up with requests, the second buffer can start serving new requests.

Once all requests in a buffer get resolved and collected, the entire buffer is discarded and a new one allocated for future requests. (Proper cycling through a ring buffer would require bookkeeping and completion handlers to collect only the known complete queries.)

In the current implementation, there is potential for a race condition when the buffer is discarded and reallocated. In practice, the race condition will never materialize so long as TracyMetalCollect() is called frequently to keep the amount of unresolved queries low.

Finally, there's a timeout mechanism during timestamp collection to detect "empty" command encoders and ensure progress.

slomp · 2024-05-17T05:37:22Z

@wolfpld I'd like to request reviews from @nosferalatu and @JamesMcCarthy44, but I can't seem to be able to add reviewers.

wolfpld · 2024-05-17T10:15:04Z

I don't know how assigning reviewers work on Github. Mentioning people should be enough to get their attention.

slomp · 2024-05-18T04:59:17Z

Also pinging @theblackunknown for a code review.

public/tracy/TracyMetal.hmm

theblackunknown

Very nice work Marcos.
I have made a first pass of code review without testing so far.
The PR description and analysis is greatly appreciated !

Can you also clarify if this code is supposed to be use with ObjC ARC or not ?
I know that codebase from codebase we can run into different expectations with this behavior.

public/tracy/TracyMetal.hmm

theblackunknown · 2024-07-25T16:53:48Z

public/tracy/TracyMetal.hmm

+        TracyMetalDebug(1<<0, TracyMetalPanic(, "MTLCounterErrorValue = 0x%llx", MTLCounterErrorValue));
+        TracyMetalDebug(1<<0, TracyMetalPanic(, "MTLCounterDontSample = 0x%llx", MTLCounterDontSample));


Do you plan on keeping all those debug points ?
Coming from the Tracy Vulkan background they look unfamiliar to me.

Yes, at least at this initial stage. Should people start reporting issues, this can help me triage the problem.

public/tracy/TracyMetal.hmm

theblackunknown · 2024-07-25T17:00:13Z

public/tracy/TracyMetal.hmm

+        TracyMetalDebug(1<<0, TracyMetalPanic(, "Calibration: CPU timestamp (Metal): %llu", cpuTimestamp));
+        TracyMetalDebug(1<<0, TracyMetalPanic(, "Calibration: GPU timestamp (Metal): %llu", gpuTimestamp));
+
+        cpuTimestamp = Profiler::GetTime();


Is this expected that you ditch the CPU timestamp returned by the MTLDevice ? Is this for consistency with other Tracy event messages ?

That's what Tracy expects. The CPU timestamp reported when creating the GPU context must be the timestamp that Tracy understands. This is consistent with other backends as well.

public/tracy/TracyMetal.hmm

theblackunknown · 2024-07-25T17:13:15Z

public/tracy/TracyMetal.hmm

+                t_start = m_mostRecentTimestamp + 5;
+                t_end = t_start + 5;


Does this means you try tp "patch" unresolved or not yet resolved timestamps ?
Can't we defer their resolution instead ?

We can't defer, because the main reason a timestamp can't be resolved is because it the command encoder with the associated timestamp(s) was never scheduled for execution.

public/tracy/TracyMetal.hmm

theblackunknown · 2024-07-25T17:21:03Z

public/tracy/TracyMetal.hmm

+    const bool m_active;
+
+    MetalCtx* m_ctx;
+    id<MTLComputeCommandEncoder> m_cmdEncoder;


This is currently unused because the above code is within #if 0 ... #endif

Yes, and I'll be working on that in a subsequent PR. I don't have hardware that supports command granularity right now.

theblackunknown · 2024-07-25T17:22:28Z

public/tracy/TracyMetal.hmm

+        MemWrite( &item->gpuZoneEnd.context, ctx->GetContextId() );
+        Profiler::QueueSerialFinish();
+
+        TracyMetalDebug(1<<2, TracyAllocN((void*)(uintptr_t)queryId, 1, "TracyMetalGpuZone"));


Are those not supposed to be pair of TracyAllocN/TracyFreeN ?

TracyFree happens in Collect, when the two timestamps announced here (begin and end timestamps) are resolved.

theblackunknown · 2024-07-25T17:23:14Z

public/tracy/TracyMetal.hmm

+private:
+    const bool m_active;
+
+    MetalCtx* m_ctx;


Alternatively you could just store the context ID

The other back-ends seem to all store the context pointer, so I'll stick to that.

slomp · 2024-05-17T17:15:22Z

public/tracy/TracyMetal.hmm

+            {
+                auto checkTime = std::chrono::high_resolution_clock::now();
+                auto requestTime = m_timestampRequestTime[k];
+                auto ms_in_flight = std::chrono::duration<float>(checkTime-requestTime).count()*1000.0f;


@wolfpld I want to remove uses of std::chrono and use what's available in Tracy already, that is, Profiler::GetTime(). I may be missing something obvious here, but how do you convert a time difference between two Profiler::GetTime() samples and convert it to, say, seconds?

Discord: https://discord.com/channels/585214693895962624/585214693895962630/1242530042366394409

public/tracy/TracyMetal.hmm

slomp · 2024-07-25T18:35:10Z

public/tracy/TracyMetal.hmm

+    const bool m_active;
+
+    MetalCtx* m_ctx;
+    id<MTLComputeCommandEncoder> m_cmdEncoder;


Yes, and I'll be working on that in a subsequent PR. I don't have hardware that supports command granularity right now.

slomp · 2024-09-10T00:41:54Z

@wolfpld Please review the changes to the manual, and if it looks good to you, this PR is ready to merge!

wolfpld · 2024-09-10T17:07:47Z

public/common/TracyQueue.hpp

@@ -401,7 +401,8 @@ enum class GpuContextType : uint8_t
    Vulkan,
    OpenCL,
    Direct3D12,
-    Direct3D11
+    Direct3D11,
+    Metal


This requires a protocol version bump (I'll do it).

nosferalatu approved these changes Jul 25, 2024

View reviewed changes

public/tracy/TracyMetal.hmm Outdated Show resolved Hide resolved

public/tracy/TracyMetal.hmm Show resolved Hide resolved

public/tracy/TracyMetal.hmm Show resolved Hide resolved

public/tracy/TracyMetal.hmm Show resolved Hide resolved

theblackunknown reviewed Jul 25, 2024

View reviewed changes

slomp commented Jul 25, 2024

View reviewed changes

Marcos Slomp added 23 commits August 26, 2024 08:15

Metal back-end WIP

1dfc926

basing metal zone scopes on MTLComputePassDescriptor

47180db

debugging timestamps...

4b695cc

fixing cpu timestamp baseline

fd1e60e

giving up on calibration, for now

94407db

fixing timestamp mapping range

3d69c48

stale comments

6d04913

collecting/resolving timestamps in pairs

c6cc424

adding blit pass and render pass interfaces

c374440

more debugging

a953879

fixing collect wrap-around

ad2c2ef

debugging

b11e665

bugfixes

04cf494

improved panic macro (supports print args)

91ca0f2

blargh

5f09d45

blarg again...

cef49c2

blarg3

638fa1f

cleanup

5ec7565

more cleanup

06b23cc

adding wiretap for debugging purposes

ea65b14

Collect pending timestamps during shutdown

799360d

comments about the decisions and behavior of the Metal back-end

aa85824

cleanup and comments

fe51f02

slomp force-pushed the slomp/metal branch from ea3610b to fe51f02 Compare August 26, 2024 16:41

Marcos Slomp added 4 commits September 3, 2024 11:15

addressing code review comments

0ffa0be

fixixng interface macros

19f376a

removing unsupported macro interfaces

bd061d1

updating documentation

c168ff6

wolfpld reviewed Sep 10, 2024

View reviewed changes

wolfpld merged commit e8ff26e into wolfpld:master Sep 10, 2024
4 checks passed

slomp mentioned this pull request Sep 13, 2024

Metal back-end slomp/tracy#1

Closed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Metal (Apple) GPU back-end for Tracy #793

Metal (Apple) GPU back-end for Tracy #793

slomp commented May 17, 2024

slomp commented May 17, 2024

wolfpld commented May 17, 2024

slomp commented May 18, 2024

theblackunknown left a comment

theblackunknown Jul 25, 2024

slomp Jul 25, 2024

theblackunknown Jul 25, 2024

slomp Jul 25, 2024

theblackunknown Jul 25, 2024

slomp Sep 3, 2024

theblackunknown Jul 25, 2024

slomp Jul 25, 2024

theblackunknown Jul 25, 2024

slomp Sep 3, 2024

theblackunknown Jul 25, 2024

slomp Sep 3, 2024

slomp May 17, 2024

slomp Jul 25, 2024

slomp Jul 25, 2024

slomp commented Sep 10, 2024

wolfpld Sep 10, 2024

		TracyMetalDebug(1<<0, TracyMetalPanic(, "MTLCounterErrorValue = 0x%llx", MTLCounterErrorValue));
		TracyMetalDebug(1<<0, TracyMetalPanic(, "MTLCounterDontSample = 0x%llx", MTLCounterDontSample));

Metal (Apple) GPU back-end for Tracy #793

Metal (Apple) GPU back-end for Tracy #793

Conversation

slomp commented May 17, 2024

slomp commented May 17, 2024

wolfpld commented May 17, 2024

slomp commented May 18, 2024

theblackunknown left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

slomp commented Sep 10, 2024

Choose a reason for hiding this comment