Control: retitle -1 gimp: crashes at start: segmentation fault in g_slice_alloc() (memory corruption?) Control: tags -1 + moreinfo
On Sun, 23 Dec 2018 at 11:24:49 +0800, 積丹尼 Dan Jacobson wrote: > Thread 1 (Thread 0x7f1aae82ce00 (LWP 3715)): > #0 0x00007f1ab0f8a484 in read () from /lib/x86_64-linux-gnu/libpthread.so.0 > No symbol table info available. > #1 0x00007f1ab1e81d34 in gimp_stack_trace_print () from > /usr/lib/libgimpbase-2.0.so.0 > No symbol table info available. > #2 0x000055bffd0fc4a0 in ?? () > No symbol table info available. > #3 0x000055bffd0fc8d8 in gimp_fatal_error () > No symbol table info available. > #4 0x000055bffd0fd037 in ?? () > No symbol table info available. > #5 <signal handler called> > No symbol table info available. > #6 0x00007f1ab118d92a in g_slice_alloc () from > /usr/lib/x86_64-linux-gnu/libglib-2.0.so.0 > No symbol table info available. > #7 0x000055bffd47ba9e in gimp_temp_buf_new () > No symbol table info available. A crash in memory allocation infrastructure is usually evidence of earlier memory corruption, so the root cause probably happened before this. I can't reproduce this crash myself. Please install debug symbols for GIMP and relevant libraries (at least libglib2.0-0-dbgsym, libgimp2.0-dbgsym and gimp-dbgsym), which you can get from the debian-debug apt repository: <https://wiki.debian.org/HowToGetABacktrace#Installing_the_debugging_symbols> This will produce a more informative backtrace. If you create a new user account on the same system and run gimp as that user, does it crash in the same way? If you run with the environment variable G_SLICE set to "debug-blocks", do you still get a crash? If you do, what is the output? If you run with environment variables G_SLICE=always-malloc and MALLOC_CHECK_=2, do you still get a crash? If you do, what is the output? Setting G_DEBUG=fatal-criticals and/or G_MESSAGES_DEBUG=all might also produce useful logging. smcv
