Hi! ----
[More or less a follow-up to my valgrind patch (see http://lists.research.att.com/pipermail/ast-developers/2013q2/002574.html) to recognise libast's allocators] The following testcase... -- snip -- $ ~/vg/bin/valgrind --num-callers=256 ./arch/linux.i386/bin/ksh -c 'set -o nounset ; bool -a a ; (( a[500][900]=1 ))' -- snip -- ... triggers the following valgrind hit with a _plain_ (e.g. no patches applied) ast-ksh.2013-05-03 on SuSE 12.3/AMD64/32bit: -- snip -- ==10072== Invalid read of size 4 ==10072== at 0x805F2D4: array_unscope (array.c:125) ==10072== by 0x8060F93: array_putval (array.c:790) ==10072== by 0x8053EA1: nv_putv (nvdisc.c:152) ==10072== by 0x809592D: nv_putval (name.c:1601) ==10072== by 0x80609CF: array_putval (array.c:649) ==10072== by 0x8053EA1: nv_putv (nvdisc.c:152) ==10072== by 0x8097E0D: _nv_unset (name.c:2529) ==10072== by 0x805BE80: nv_settype (nvtype.c:1460) ==10072== by 0x80DD4CC: setall (typeset.c:726) ==10072== by 0x80DC511: b_typeset (typeset.c:455) ==10072== by 0x80B7888: sh_exec (xec.c:1357) ==10072== by 0x80BA453: sh_exec (xec.c:2218) ==10072== by 0x805337E: exfile (main.c:588) ==10072== by 0x8052725: sh_main (main.c:360) ==10072== by 0x805149A: main (pmain.c:45) ==10072== Address 0x42a5940 is 40 bytes inside a block of size 224 free'd ==10072== at 0x402A29F: _ast_free (vg_replace_malloc.c:1001) ==10072== by 0x8060ED0: array_putval (array.c:777) ==10072== by 0x8053EA1: nv_putv (nvdisc.c:152) ==10072== by 0x809592D: nv_putval (name.c:1601) ==10072== by 0x80609CF: array_putval (array.c:649) ==10072== by 0x8053EA1: nv_putv (nvdisc.c:152) ==10072== by 0x8097E0D: _nv_unset (name.c:2529) ==10072== by 0x805BE80: nv_settype (nvtype.c:1460) ==10072== by 0x80DD4CC: setall (typeset.c:726) ==10072== by 0x80DC511: b_typeset (typeset.c:455) ==10072== by 0x80B7888: sh_exec (xec.c:1357) ==10072== by 0x80BA453: sh_exec (xec.c:2218) ==10072== by 0x805337E: exfile (main.c:588) ==10072== by 0x8052725: sh_main (main.c:360) ==10072== by 0x805149A: main (pmain.c:45) -- snip -- David: Can you verify whether this is a real bug, please (I'd like to verify that my patch to valgrind is correct) ? ---- Bye, Roland -- __ . . __ (o.\ \/ /.o) [email protected] \__\/\/__/ MPEG specialist, C&&JAVA&&Sun&&Unix programmer /O /==\ O\ TEL +49 641 3992797 (;O/ \/ \O;) _______________________________________________ ast-developers mailing list [email protected] http://lists.research.att.com/mailman/listinfo/ast-developers
