diff options
author | Austin Yuan <shengquan.yuan@gmail.com> | 2010-10-26 10:18:11 -0700 |
---|---|---|
committer | Austin Yuan <shengquan.yuan@gmail.com> | 2010-10-26 10:18:11 -0700 |
commit | 866360c82ea76182bfa601a074c87d4c83325419 (patch) | |
tree | bcdf75eea76d17a67707625cccdc9d3803cd98cb /va/va_fool.c | |
parent | fe9ee747b973118f21a9b838dd02972b96924c1e (diff) | |
download | libva-866360c82ea76182bfa601a074c87d4c83325419.tar.gz |
more trace for displayattribute/vaPutSurface
Signed-off-by: Austin Yuan <shengquan.yuan@gmail.com>
Diffstat (limited to 'va/va_fool.c')
-rw-r--r-- | va/va_fool.c | 14 |
1 files changed, 10 insertions, 4 deletions
diff --git a/va/va_fool.c b/va/va_fool.c index f4d866f..bdb2fb9 100644 --- a/va/va_fool.c +++ b/va/va_fool.c @@ -52,28 +52,31 @@ * . if set, encode does nothing, but fill in a hard-coded 720P clip into coded buffer. * . VA CONTEXT/CONFIG/SURFACE will call into drivers, but VA Buffer creation is done in here * . Bypass all ~SvaBeginPic/vaRenderPic/vaEndPic~T + * LIBVA_FOOL_POSTP: + * . if set, do nothing for vaPutSurface */ /* global settings */ -/* LIBVA_FOOL_DECODE/LIBVA_FOOL_ENCODE */ +/* LIBVA_FOOL_DECODE/LIBVA_FOOL_ENCODE/LIBVA_FOOL_POSTP */ static int fool_decode = 0; static int fool_encode = 0; +int fool_postp = 0; #define FOOL_CONTEXT_MAX 4 /* per context settings */ static struct _fool_context { - VADisplay dpy; + VADisplay dpy; /* should use context as the key */ VAProfile fool_profile; /* current profile for buffers */ VAEntrypoint fool_entrypoint; /* current entrypoint */ FILE *fool_fp_codedclip; /* load a clip from disk for fooling encode*/ - /* all buffers with same type share one memory + /* all buffers with same type share one malloc-ed memory * bufferID = (buffer numbers with the same type << 8) || type - * the buffer memory can be find by fool_buf[bufferID & 0xff] + * the malloc-ed memory can be find by fool_buf[bufferID & 0xff] * the size is ignored here */ char *fool_buf[VABufferTypeMax]; /* memory of fool buffers */ @@ -146,6 +149,9 @@ void va_FoolInit(VADisplay dpy) if (fool_index == FOOL_CONTEXT_MAX) return; + if (va_parseConfig("LIBVA_FOOL_POSTP", NULL) == 0) + fool_postp = 1; + if (va_parseConfig("LIBVA_FOOL_DECODE", NULL) == 0) fool_decode = 1; |