On Thu, 06 Feb 2003 07:46:35 +1000
Chris Ison <[EMAIL PROTECTED]> wrote:
> a couple of us QF codes are questioning this, looks like gcc
> optimizations or gdb are interefering cause the way QF handles the
> vertex array (with error checking), it's impossable for starters for it
> send a NULL indices pointer, not to mention a 0 count.
Looks like you're right. I recompiled the radeon driver with -O0 and got
this:
#0 0x40334581 in kill () from /lib/libc.so.6
#1 0x40334394 in raise () from /lib/libc.so.6
#2 0x403358d1 in abort () from /lib/libc.so.6
#3 0x4032edb2 in __assert_fail () from /lib/libc.so.6
#4 0x41654522 in radeonAllocDmaRegion (rmesa=0x8164680, region=0x81670c0,
bytes=72000, alignment=3) at radeon_ioctl.c:635
#5 0x4165455e in radeonAllocDmaRegionVerts (rmesa=0x8164680,
region=0x81670c0, numverts=3000, vertsize=24, alignment=4)
at radeon_ioctl.c:647
#6 0x4165996a in radeonEmitArrays (ctx=0x8167898, inputs=265)
at radeon_maos_verts.c:311
#7 0x4167d063 in radeon_run_tcl_render (ctx=0x8167898, stage=0x82e2d38)
at radeon_tcl.c:318
#8 0x4162e6b3 in _tnl_run_pipeline (ctx=0x8167898) at t_pipeline.c:154
#9 0x41666a9b in radeonWrapRunPipeline (ctx=0x8167898) at radeon_state.c:2088
#10 0x4161f095 in _tnl_DrawElements (mode=2823, count=3000, type=137243856,
indices=0x8318bb8) at t_array_api.c:99
#11 0x415ba8b6 in neutral_DrawElements (mode=0, count=0, type=0, indices=0x0)
at ../../../../extras/Mesa/src/vtxfmt_tmp.h:369
#12 0x41685fde in radeon_fallback_DrawElements (mode=7, count=3000, type=5125,
indices=0x8318bb8) at ../../../../../../extras/Mesa/src/vtxfmt_tmp.h:369
#13 0x415ba8b6 in neutral_DrawElements (mode=0, count=0, type=0, indices=0x0)
at ../../../../extras/Mesa/src/vtxfmt_tmp.h:369
---Type <return> to continue, or q <return> to quit---
#14 0x40026f38 in Draw_nString (x=6, y=1078070736,
str=0x4144e780 "_histogram GL_EXT_packed_pixels GL_EXT_polygon_offset", ' '
<repeats 19 times>, "GL_EXT_rescale_normal GL_EXT_secondary_color GL_EXT_texture3D", '
' <repeats 17 times>, "GL_EXT_texture_env_add GL_EXT_texture_env_combine "...,
count=0) at gl_draw.c:367
#15 0x72676f74 in ?? ()
Cannot access memory at address 0x7369685f
There are still a few 0 counts but the source of neutral_DrawElements is
in a different directory and is still compiled with -O3. The 3000 looks
more sensible. With vertex size 24 this makes 72000 bytes which is more
than fits into one 64KB DMA buffer.
Felix
__\|/__ ___ ___ ___
__Tschüß_______\_6 6_/___/__ \___/__ \___/___\___You can do anything,___
_____Felix_______\Ä/\ \_____\ \_____\ \______U___just not everything____
[EMAIL PROTECTED] >o<__/ \___/ \___/ at the same time!
-------------------------------------------------------
This SF.NET email is sponsored by:
SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See!
http://www.vasoftware.com
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
