This is the bug/patch tracker for Rockbox. Click here for more information.
Quick links: Bugs · Patches · Rockbox frontpage
FS#10814 - Crossfade crashes simulator
Attached to Project:
Rockbox
Opened by Steve Bavin (pondlife) - Thursday, 26 November 2009, 13:30 GMT+2
Last edited by Steve Bavin (pondlife) - Thursday, 26 November 2009, 15:41 GMT+2
Opened by Steve Bavin (pondlife) - Thursday, 26 November 2009, 13:30 GMT+2
Last edited by Steve Bavin (pondlife) - Thursday, 26 November 2009, 15:41 GMT+2
|
DetailsUsing the attached settings, an H300 simulator (r23757)
gdb gives the following output and backtrace:: Program received signal SIGSEGV, Segmentation fault. [Switching to thread 7016.0x113c] 0x004498dd in crossfade_mix_fade (factor=0, length=0, buf=0x0, out_sample=0x470fca8, out_chunk=0x470fca4) at /home/Steve/rockbox/apps/pcmbuf.c:726 726 int16_t *output_buf = (int16_t *)((*out_chunk)->addr); (gdb) bt #0 0x004498dd in crossfade_mix_fade (factor=0, length=0, buf=0x0, out_sample=0x470fca8, out_chunk=0x470fca4) at /home/Steve/rockbox/apps/pcmbuf.c:726 #1 0x00449c01 in crossfade_start () at /home/Steve/rockbox/apps/pcmbuf.c:857 #2 0x0044932d in pcmbuf_request_buffer (count=0x470fd08) at /home/Steve/rockbox/apps/pcmbuf.c:374 #3 0x0044a640 in codec_pcmbuf_insert_callback (ch1=0x10019ad4, ch2=0x1001acd4, count=623) at /home/Steve/rockbox/apps/codec_thread.c:186 #4 0x10001b7c in codec_main () at /home/Steve/rockbox/apps/codecs/mpa.c:456 #5 0x10001d77 in codec_start () at /home/Steve/rockbox/apps/codecs/codec_crt0.c:50 #6 0x0044d6a7 in codec_load_ram (size=222865, api=0x47d060) at /home/Steve/rockbox/apps/codecs.c:232 #7 0x0044d813 in codec_load_file (plugin=0x488160 "mpa", api=0x47d060) at /home/Steve/rockbox/apps/codecs.c:273 #8 0x0044ad42 in codec_thread () at /home/Steve/rockbox/apps/codec_thread.c:492 #9 0x00464d2a in runthread (data=0x6c9ca0) at /home/Steve/rockbox/uisimulator/sdl/thread-sdl.c:468 #10 0x6bfb0adc in SDL_RunThread (data=0x30d3d50) at ../../../SDL-1.2.9/src/thread/SDL_thread.c:218 #11 0x6bfb0dd1 in RunThread (data=0x30d3d50) at SDL_systhread.c:44 #12 0x7c80b729 in KERNEL32!GetModuleFileNameA () from /cygdrive/c/WINDOWS/system32/kernel32.dll #13 0x00000000 in ?? () |
This task depends upon
Closed by Steve Bavin (pondlife)
Thursday, 26 November 2009, 15:41 GMT+2
Reason for closing: Fixed
Additional comments about closing: Should be fine now.
Thursday, 26 November 2009, 15:41 GMT+2
Reason for closing: Fixed
Additional comments about closing: Should be fine now.
I should add that this only happens on natural track change, not manal skip.
More info - it was introduced in r23628.
Here's a patch that fixes it, but I'm not sure if this is the correct fix or just a sticking plaster.
My apologies. I thought that function tolerated a zero length input, but if it doesn't, this fix won't hurt anything. I'd go ahead and commit the patch. The whole function will probably be replaced in the not too distant future anyway so if the patch prevents crashes now, it's useful.