Opened 11 years ago

Closed 11 years ago

Last modified 11 years ago

#2729 closed defect (fixed)

[raster]: raster crashes when compiled with --enable-debug under windows

Reported by: robe Owned by: dustymugs
Priority: blocker Milestone: PostGIS 2.1.3
Component: raster Version: 2.1.x
Keywords: windows Cc:

Description

I mentioned the issue in #2726

I isoloated the offending line to

rt_pg.c line 112

POSTGIS_RT_DEBUGF(4, "GDAL_SKIP = %s", CPLGetConfigOption("GDAL_SKIP", NULL));

so I remarked it out. This might be just an issue under windows. I don't have GDAL_SKIP defined (I have to recheck what happens if I do) and I think windows has issue with null strings so that might be the cause.

Anyway not quite sure what to replace that with.

Change History (6)

comment:1 by strk, 11 years ago

Copying the stack trace over:

#0  0x000007fefd8e5a01 in strlen () from C:\Windows\system32\msvcrt.dll
#1  0x000000000074df39 in fmtstr (target=0x23ce320, pointflag=<optimized out>, maxwidth=0, minlen=0, leftjust=0, value=0x0) at snprintf.c:779
#2  dopr (target=target@entry=0x23ce320, format=0x39b131 "", args=0x23ce458 "A4ñü_\a", args@entry=0x23ce438 "") at snprintf.c:714
#3  0x000000000074e13f in pg_vsnprintf (str=<optimized out>, count=count@entry=1023, fmt=fmt@entry=0x39b118 "[%s:%s:%d] GDAL_SKIP = %s", args=args@entry=0x23ce438 "") at snprintf.c:154
#4  0x000000000057c0b1 in appendStringInfoVA (str=str@entry=0x23ce3e0, fmt=fmt@entry=0x39b118 "[%s:%s:%d] GDAL_SKIP = %s", args=args@entry=0x23ce438 "") at stringinfo.c:137
#5  0x000000000071607f in errmsg_internal (fmt=fmt@entry=0x68701c5f "[%s:%s:%d] GDAL_SKIP = %s") at elog.c:821
#6  0x0000000068684826 in rtpg_assignHookGDALEnabledDrivers () at rt_pg.c:112
#7  _PG_init () at rt_pg.c:237

comment:2 by robe, 11 years ago

Resolution: fixed
Status: newclosed

the null patching seems to have fixed this.

So as of r12521 no longer getting crashing.

comment:3 by robe, 11 years ago

Resolution: fixed
Status: closedreopened

comment:4 by robe, 11 years ago

actually let me check again. I realized I must be running an older version since that doesn't match your new.

comment:5 by robe, 11 years ago

Resolution: fixed
Status: reopenedclosed

okay it is fixed as of r12523

comment:6 by robe, 11 years ago

for record I think the fix was in r12522

Note: See TracTickets for help on using tickets.