There are two usage types of buffer CPU accesses:
One where we try to use the buffer contents for multiple draw commands in
a batch. (batch := sequence of commands that are flushed together),
like incrementally adding bitmaps to a bitmap texture that is reallocated
on flush.
And one where we assume we can safely overwrite the old buffer contexts, like
glTexSubImage. In this case we need to make sure all old drawing commands
referencing the buffer are flushed before we map the buffer.
This is easily forgotten.
Add wrappers for the most common of these operations. The first type is
prefixed with "st_no_flush" and the second type is prefixed with
"st_cond_flush", where "cond" indicates that we attmpt to only flush
if there is indeed unflushed draw commands referencing the buffer.
Prefixed functions are
screen::get_tex_transfer
pipe_buffer_write
pipe_buffer_read
pipe_buffer_map
Please use the wrappers whenever possible.
Signed-off-by: Thomas Hellstrom <thellstrom-at-vmware-dot-com>
Use proper fields for marking if fp is translated, and if is translated succesfully.
Now if fp gets translated (even unsuccesfully) fp->translated is true. If the translation failed (i.e. because we exceeded limit of
maximum texture indirections) the fp->error is set. With a little updated fallback function it prevents non native fragment programs
from beeing translated with every frame (the translation would fail anyway so there's no point to try again).
Also implement IsProgramNative function for GL_FRAGMENT_PROGRAM_ARB (it should give some performance boost in apps that checks if
program is native and falls back to simpler shader to meet hw limits if necessary) and cleanup indentation (remove whitespaces on empty
lines).
Also enable them all regardless of screen bpp, as 32 bpp what I've been
testing against, and haven't been able to detect any screen bpp-specific
troubles with them.
For some reason, MOV instructions using immediate src values don't seem
to work reliably on the GLSL path. Disable them for now (falling back to
const buffer reads). This fixes a bunch of glean glsl1 failures.
need to round up height for _mesa_copy_rect otherwise
textures with height smaller than 4 won't get copied to the miptree at all
Also fix up the confusing debug output (don't output unitialized values,
and output if data is present and the compressed flag)
valgrind was showing a race between the drawable getting destroyed
by the X resource freeing code, and the context getting destroyed
later and freeing the drawable.
However I've no idea if some other combination of things could cause
this code to leak.
Also implement context member functions to optimize away those
flushes whenever possible.
Signed-off-by: Thomas Hellstrom <thellstrom-at-vmware-dot-com>
