Here I step into FS_NativePath and see that relativeto is set to FS_BINARYPATH. Eventually out is written missing slash in path: /usr/local/libexec/qwfteplug_ezhud_amd64.so (gdb) run Starting program: /usr/local/libexec/qw/fteqw -basedir ~/games/data/quake [Switching to thread 1 (thread 128644)](running) [New thread 310335] [New thread 572420] [New thread 179982] [New thread 426655] Thread 1 hit Breakpoint 4, Plug_Load (file=0xf92404ab0cc "fteplug_ezhud_amd64.so") at common/plugin.c:209 209 if (!Q_strncasecmp(file,...
Sorry for the broken formatting. Here I repaste solution that worked for me, but I am unsure if it will break other filesystem paths: solution: $OpenBSD$ Index: engine/common/fs.c --- engine/common/fs.c.orig +++ engine/common/fs.c @@ -2058,7 +2058,7 @@ qboolean FS_NativePath(const char fname, enum fs_rela if (host_parms.binarydir && host_parms.binarydir) snprintf(out, outlen, "%s%s", host_parms.binarydir, fname); else - snprintf(out, outlen, "%s%s", host_parms.basedir, fname); + snprintf(out, outlen,...
FS_NativePath problems in loading plugin OpenBSD
segfault at end of replay on OpenBSD
Reverting to revision 5554 implementation of Font_LineBreaks, as in the following patch, seems to have resolved the issue. $OpenBSD$ revert Font_LineBreaks implementation from revision 5675 to 5554 implementation to resolve frequent segfaults https://sourceforge.net/p/fteqw/tickets/80/ Index: engine/gl/gl_font.c --- engine/gl/gl_font.c.orig +++ engine/gl/gl_font.c @@ -2665,10 +2665,7 @@ int Font_LineBreaks(conchar_t *start, conchar_t *end, if (codepoint > ' ') l = n; else - { - l = n; break; - }...
segfault FontDecode and Font_LineBreaks on OpenBSD
SDL2 keybinds
100 fps OpenBSD SDL2 build
Resolved with revision 5506. https://sourceforge.net/p/fteqw/code/5506/ I received help in #freecs. Spoike suggested the fix and eukara implemented it. To reproduce on the older revision 5505, I repeatedly shot the glock, switched to and from burst fire, dropped the glock and spammed 'flush' (with the up arrow for last command) in console, per Spoike's suggestion. This would cause the same segfault in the OP to occur within several minutes. Revision 5506 seems to not exhibit this behavior. 22:51...
FreeCS segfault