From: Pauli Nieminen Date: Tue, 18 Aug 2009 15:51:38 +0000 (+0300) Subject: libdrm_radeon: Optimize copy of table to cs buffer with specialized call. X-Git-Tag: submit/1.0/20121108.012404~433 X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=8c43b79b21929e9e54e13e892f7787e222e73f39;p=profile%2Fivi%2Flibdrm.git libdrm_radeon: Optimize copy of table to cs buffer with specialized call. Using this call in OUT_BATCH_TABLE reduces radeonEmitState cpu usage from 9% to 5% and emit_vpu goes from 7% to 1.5%. I did use calgrind to profile gears for cpu hotspots with r500 card. Signed-off-by: Pauli Nieminen --- diff --git a/libdrm/radeon/radeon_cs.h b/libdrm/radeon/radeon_cs.h index 7efec7e..1117a85 100644 --- a/libdrm/radeon/radeon_cs.h +++ b/libdrm/radeon/radeon_cs.h @@ -201,6 +201,15 @@ static inline void radeon_cs_write_qword(struct radeon_cs *cs, uint64_t qword) } } +static inline void radeon_cs_write_table(struct radeon_cs *cs, void *data, uint32_t size) +{ + memcpy(cs->packets + cs->cdw, data, size * 4); + cs->cdw += size; + if (cs->section) { + cs->section_cdw += size; + } +} + static inline void radeon_cs_space_set_flush(struct radeon_cs *cs, void (*fn)(void *), void *data) { cs->space_flush_fn = fn;