bugzilla-daemon at mindrot.org
2002-Dec-18 16:16 UTC
[Bug 458] sshd crashes with "fatal: mm_malloc: size too big"
http://bugzilla.mindrot.org/show_bug.cgi?id=458 ------- Additional Comments From us- at gmx.de 2002-12-19 03:16 ------- Sorry, typo: The irix version is 6.5.17 not 6.5.18 ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.
bugzilla-daemon at mindrot.org
2002-Dec-30 21:57 UTC
[Bug 458] sshd crashes with "fatal: mm_malloc: size too big"
http://bugzilla.mindrot.org/show_bug.cgi?id=458 groeskens at bluewin.ch changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |groeskens at bluewin.ch ------- Additional Comments From groeskens at bluewin.ch 2002-12-31 08:57 ------- Same Problem on Solaris 8 Build 32-Bit, gcc 2.95-3 It seems that line 170 in monitor_mm.c is the problem: if (size > SIZE_T_MAX - MM_MINSIZE + 1) fatal("mm_malloc: size too big"); size seems to be a different type (uint32) in defines.h: --- #ifndef SIZE_T_MAX #define SIZE_T_MAX ULONG_MAX #endif /* SIZE_T_MAX */ #ifndef HAVE_SIZE_T typedef unsigned int size_t; # define HAVE_SIZE_T #endif /* HAVE_SIZE_T */ --- the line #define SIZE_T_MAX ULONG_MAX should be changed, I tried # define SIZE_T_MAX ((2 << (8 * sizeof(size_t)) - 1) but got an error (also out of bounds :-( The definition of SIZE_T_MAX has to be changed to be in the range of the type of size From the build log: --- gcc -O3 -pipe -mcpu=ultrasparc -m32 -I. -I. -I/usr/local/include -I/usr/local/i nclude -DSSHDIR=\"/etc/ssh\" -D_PATH_SSH_PROGRAM=\"/usr/local/bin/ssh\" -D_PATH_ SSH_ASKPASS_DEFAULT=\"/usr/local/lib/ssh/ssh-askpass\" -D_PATH_SFTP_SERVER=\"/us r/local/lib/ssh/sftp-server\" -D_PATH_SSH_KEY_SIGN=\"/usr/local/lib/ssh/ssh-keys ign\" -D_PATH_SSH_PIDDIR=\"/var/run\" -D_PATH_PRIVSEP_CHROOT_DIR=\"/var/empty\" -DSSH_RAND_HELPER=\"/usr/local/lib/ssh/ssh-rand-helper\" -DHAVE_CONFIG_H -c monitor_mm.c monitor_mm.c: In function `mm_malloc': monitor_mm.c:170: warning: integer overflow in expression monitor_mm.c:170: warning: comparison is always true due to limited range of data type --- -- Guido ------- You are receiving this mail because: ------- You are the assignee for the bug, or are watching the assignee.