x86: Fix the test for negative pixel count in optimized rgb565 spans.

There's a bunch of bogus looking stuff the count handling in this code, but
this fixes the testcases we have.
This commit is contained in:
Eric Anholt 2009-10-30 17:22:58 -07:00
parent 21a3a79371
commit d63c29ef20

View file

@ -522,7 +522,7 @@ _generic_read_RGBA_span_RGB565_MMX:
addl $32,%esp
sarl $2, %ecx
jle .L01 /* Bail early if the count is negative. */
jl .L01 /* Bail early if the count is negative. */
jmp .L02
.L03: