Stuart Henderson a écrit :
On 2009/09/05 16:50, frantisek holop wrote:
is it just my crazy system, or is the executable really ~150M?
ah, thanks for spotting, looks like it's building with symbols and
not stripping (which takes it down to a more reasonable 3.3M).
it also dumps core on my eeepc, after any kind of
error, for example:
Warning: [Main]: mixxx: cannot connect to X server localhost:10.0
Segmentation fault (core dumped)
or after cancelling out the music library chooser dialog, i get:
Failed to initilaize GEM. Falling back to classic.
Segmentation fault (core dumped)
rebuilding lib/libGL with this diff from oga@ should fix the segfaults.
Index: intel_context.c
===================================================================
RCS file: /cvs/xenocara/dist/Mesa/src/mesa/drivers/dri/intel/intel_context.c,v
retrieving revision 1.7
diff -u -p -r1.7 intel_context.c
--- intel_context.c 6 Aug 2009 15:50:02 -0000 1.7
+++ intel_context.c 5 Sep 2009 15:58:37 -0000
@@ -926,39 +926,6 @@ intelDestroyContext(__DRIcontextPrivate
*/
}
- /* XXX In intelMakeCurrent() below, the context's static regions are
- * referenced inside the frame buffer; it's listed as a hack,
- * with a comment of "XXX FBO temporary fix-ups!", but
- * as long as it's there, we should release the regions here.
- * The do/while loop around the block is used to allow the
- * "continue" statements inside the block to exit the block,
- * to avoid many layers of "if" constructs.
- */
- do {
- __DRIdrawablePrivate * driDrawPriv = intel->driDrawable;
- struct intel_framebuffer *intel_fb;
- struct intel_renderbuffer *irbDepth, *irbStencil;
- if (!driDrawPriv) {
- /* We're already detached from the drawable; exit this block. */
- continue;
- }
- intel_fb = (struct intel_framebuffer *) driDrawPriv->driverPrivate;
- if (!intel_fb) {
- /* The frame buffer is already gone; exit this block. */
- continue;
- }
- irbDepth = intel_get_renderbuffer(&intel_fb->Base, BUFFER_DEPTH);
- irbStencil = intel_get_renderbuffer(&intel_fb->Base, BUFFER_STENCIL);
-
- /* Usually, the stencil buffer is the same as the depth buffer;
- * but they're handled separately in MakeCurrent, so we'll
- * handle them separately here.
- */
- if (irbStencil && irbStencil->region == intel->depth_region) {
- intel_renderbuffer_set_region(irbStencil, NULL);
- }
- } while (0);
-
intel_region_release(&intel->front_region);
intel_region_release(&intel->back_region);
intel_region_release(&intel->depth_region);
Hi,
Not usable on my amd64 laptop (intel 945 gms), but works fine on my i386
desktop (P4 3Ghz, ati x1650).
I have the same strip problem :
16:16:26] d...@desktop:~ $ ls -lh /usr/local/bin/mixxx
-rwxr-xr-x 1 root wheel 136M Sep 26 16:06 /usr/local/bin/mixxx
Thanks for this port !
Btw any news about lmms ?
