Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(13)

Unified Diff: source/row_neon.cc

Issue 2478313004: HalfFloat neon armv7 fix for destination pointer. (Closed)
Patch Set: added caveat on rounding difference Created 4 years, 1 month ago
Use n/p to move between diff chunks; N/P to move between comments. Draft comments are only viewable by you.
Jump to:
View side-by-side diff with in-line comments
Download patch
« no previous file with comments | « include/libyuv/version.h ('k') | source/row_neon64.cc » ('j') | no next file with comments »
Expand Comments ('e') | Collapse Comments ('c') | Show Comments Hide Comments ('s')
Index: source/row_neon.cc
diff --git a/source/row_neon.cc b/source/row_neon.cc
index 9385b275dc4253bcd7c4830f3b44145ff094e724..c31fdcd6f082c3062ed9cd49dc60653955e851e0 100644
--- a/source/row_neon.cc
+++ b/source/row_neon.cc
@@ -2742,16 +2742,16 @@ void HalfFloat1Row_NEON(const uint16* src, uint16* dst, float, int width) {
MEMACCESS(0)
"vld1.8 {q1}, [%0]! \n" // load 8 shorts
"subs %2, %2, #8 \n" // 8 pixels per loop
- "vmovl.u8 q2, d2 \n" // 8 int's
- "vmovl.u8 q3, d3 \n"
+ "vmovl.u16 q2, d2 \n" // 8 int's
+ "vmovl.u16 q3, d3 \n"
"vcvt.f32.u32 q2, q2 \n" // 8 floats
"vcvt.f32.u32 q3, q3 \n"
"vmul.f32 q2, q2, q0 \n" // adjust exponent
"vmul.f32 q3, q3, q0 \n"
- "vqshrn.u32 d2, q2, #13 \n" // isolate halffloat
+ "vqshrn.u32 d2, q2, #13 \n" // isolate halffloat
"vqshrn.u32 d3, q3, #13 \n"
MEMACCESS(1)
- "vst1.8 {q1}, [%0]! \n"
+ "vst1.8 {q1}, [%1]! \n"
"bgt 1b \n"
: "+r"(src), // %0
"+r"(dst), // %1
@@ -2770,16 +2770,16 @@ void HalfFloatRow_NEON(const uint16* src, uint16* dst, float scale, int width) {
MEMACCESS(0)
"vld1.8 {q1}, [%0]! \n" // load 8 shorts
"subs %2, %2, #8 \n" // 8 pixels per loop
- "vmovl.u8 q2, d2 \n" // 8 int's
- "vmovl.u8 q3, d3 \n"
+ "vmovl.u16 q2, d2 \n" // 8 int's
+ "vmovl.u16 q3, d3 \n"
"vcvt.f32.u32 q2, q2 \n" // 8 floats
"vcvt.f32.u32 q3, q3 \n"
"vmul.f32 q2, q2, q0 \n" // adjust exponent
"vmul.f32 q3, q3, q0 \n"
- "vqshrn.u32 d2, q2, #13 \n" // isolate halffloat
+ "vqshrn.u32 d2, q2, #13 \n" // isolate halffloat
"vqshrn.u32 d3, q3, #13 \n"
MEMACCESS(1)
- "vst1.8 {q1}, [%0]! \n"
+ "vst1.8 {q1}, [%1]! \n"
"bgt 1b \n"
: "+r"(src), // %0
"+r"(dst), // %1
« no previous file with comments | « include/libyuv/version.h ('k') | source/row_neon64.cc » ('j') | no next file with comments »

Powered by Google App Engine
This is Rietveld 408576698