Commit 9e0f9464 authored by Chris Wilson's avatar Chris Wilson

drm/i915/gem: Async GPU relocations only

Reduce the 3 relocation paths down to the single path that accommodates
all. The primary motivation for this is to guard the relocations with a
natural fence (derived from the i915_request used to write the
relocation from the GPU).

The tradeoff in using async gpu relocations is that it increases latency
over using direct CPU relocations, for the cases where the target is
idle and accessible by the CPU. The benefit is greatly reduced lock
contention and improved concurrency by pipelining.

Note that forcing the async gpu relocations does reveal a few issues
they have. Firstly, is that they are visible as writes to gem_busy,
causing to mark some buffers are being to written to by the GPU even
though userspace only reads. Secondly is that, in combination with the
cmdparser, they can cause priority inversions. This should be the case
where the work is being put into a common workqueue losing our priority
information and so being executed in FIFO from the worker, denying us
the opportunity to reorder the requests afterwards.
Signed-off-by: default avatarChris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: default avatarMatthew Auld <matthew.auld@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200604211457.19696-1-chris@chris-wilson.co.uk
parent 2d387995
...@@ -37,20 +37,14 @@ static int __igt_gpu_reloc(struct i915_execbuffer *eb, ...@@ -37,20 +37,14 @@ static int __igt_gpu_reloc(struct i915_execbuffer *eb,
return err; return err;
/* 8-Byte aligned */ /* 8-Byte aligned */
if (!__reloc_entry_gpu(eb, vma, err = __reloc_entry_gpu(eb, vma, offsets[0] * sizeof(u32), 0);
offsets[0] * sizeof(u32), if (err)
0)) {
err = -EIO;
goto unpin_vma; goto unpin_vma;
}
/* !8-Byte aligned */ /* !8-Byte aligned */
if (!__reloc_entry_gpu(eb, vma, err = __reloc_entry_gpu(eb, vma, offsets[1] * sizeof(u32), 1);
offsets[1] * sizeof(u32), if (err)
1)) {
err = -EIO;
goto unpin_vma; goto unpin_vma;
}
/* Skip to the end of the cmd page */ /* Skip to the end of the cmd page */
i = PAGE_SIZE / sizeof(u32) - RELOC_TAIL - 1; i = PAGE_SIZE / sizeof(u32) - RELOC_TAIL - 1;
...@@ -60,12 +54,9 @@ static int __igt_gpu_reloc(struct i915_execbuffer *eb, ...@@ -60,12 +54,9 @@ static int __igt_gpu_reloc(struct i915_execbuffer *eb,
eb->reloc_cache.rq_size += i; eb->reloc_cache.rq_size += i;
/* Force batch chaining */ /* Force batch chaining */
if (!__reloc_entry_gpu(eb, vma, err = __reloc_entry_gpu(eb, vma, offsets[2] * sizeof(u32), 2);
offsets[2] * sizeof(u32), if (err)
2)) {
err = -EIO;
goto unpin_vma; goto unpin_vma;
}
GEM_BUG_ON(!eb->reloc_cache.rq); GEM_BUG_ON(!eb->reloc_cache.rq);
rq = i915_request_get(eb->reloc_cache.rq); rq = i915_request_get(eb->reloc_cache.rq);
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment