From: Chris Wilson Date: Fri, 21 Jul 2017 15:36:48 +0000 (+0100) Subject: i965: Ignore reloc read/write domains X-Git-Url: https://git.libre-soc.org/?a=commitdiff_plain;h=ba9b71e56a43f95946dbb0329f99f2db098005b0;p=mesa.git i965: Ignore reloc read/write domains Since before the kernel supported I915_EXEC_NO_RELOC, long before our minimum kernel requirement, the kernel unconditionally invalidated all GPU TLBs before a batch and flushed all GPU caches after a batch. At that moment, the only use for read/write domain was for activity tracking, ensuring that future reads waited for the last writer and future writes waited for all reads. This only requires a single bit in the execbuf interface which can be supplied via the NO_RELOC interface, making the use of relocation domains entirely redundant. Trimming the excess writes into the array allows the compiler to be much more frugal: text data bss dec hex filename 8493790 357184 424944 9275918 8d8a0e i965_dri.baseline 8493758 357184 424944 9275886 8d89ee i965_dri.so (This text improvement really does come from dropping domains, not from the new use of C99 initializers.) Signed-off-by: Chris Wilson Reviewed-by: Kenneth Graunke --- diff --git a/src/mesa/drivers/dri/i965/intel_batchbuffer.c b/src/mesa/drivers/dri/i965/intel_batchbuffer.c index dcc6f90238c..672a960868f 100644 --- a/src/mesa/drivers/dri/i965/intel_batchbuffer.c +++ b/src/mesa/drivers/dri/i965/intel_batchbuffer.c @@ -811,17 +811,13 @@ brw_emit_reloc(struct intel_batchbuffer *batch, uint32_t batch_offset, offset64 = target->offset64; } - struct drm_i915_gem_relocation_entry *reloc = - &batch->relocs[batch->reloc_count]; - - batch->reloc_count++; - - reloc->offset = batch_offset; - reloc->delta = target_offset; - reloc->target_handle = target->gem_handle; - reloc->read_domains = read_domains; - reloc->write_domain = write_domain; - reloc->presumed_offset = offset64; + batch->relocs[batch->reloc_count++] = + (struct drm_i915_gem_relocation_entry) { + .offset = batch_offset, + .delta = target_offset, + .target_handle = target->gem_handle, + .presumed_offset = offset64, + }; /* Using the old buffer offset, write in what the right data would be, in * case the buffer doesn't move and we can short-circuit the relocation