Bus error when setting antispamdomain directive

[jkbzh]

I’m been playing with the MacPorts version of the hypermail 2.3.0 release. I have discovered that assigning any value to the antispamdomain directive (quoted, unquoted, whatever) results in a bus error via stdout. This error did not occur with the 2.2.0 release.

Please let me know if there is any additional information you require.

Reported by: splorp

[jkbzh]

Hi, could you try applying the following patch and see if it fixes your problem? Thanks.
(I would have attached the file if I knew how to do that!)
-j

[[[[ patch starts here below, remove this line ]]]
*** getname.c.orig 2009-12-11 15:15:07.000000000 +0100
--- getname.c 2009-12-11 15:18:09.000000000 +0100

---

*** 63,69 ****
}

if (atptr) {
! char newbuf = malloc(newlen);
int index = atptr - input;
/ copy the part before the @ */
memcpy(newbuf, input, index);
--- 63,69 ----
}

if (atptr) {
! char newbuf = malloc(newlen + 1);
int index = atptr - input;
/ copy the part before the @ */
memcpy(newbuf, input, index);

Original comment by: jkbzh

[jkbzh]

patch to fix the bus dump

Original comment by: jkbzh

[jkbzh]

Thanks for the patch. I’ve applied this manually, but discovered that the MacPorts build is not applying other related patches to files within subdirectories (such as src/). I need to manually apply the remaining MacPorts patches to a clean copy of the 2.3.0 source and will report back.

Original comment by: splorp

[jkbzh]

I've built hypermail 2.3.0 (w/patch.diff applied) & pcre 8.0.0 from scratch for Mac OS X 10.5.3 Server (running on a Core 2 Duo) and I can still reproduce this bus error ('10Bus error') during loading the mbox file.

Note: I'm working with two mbox files out of a large set: one mbox file (2MB) does not produce the bus error, but the other (1.2MB) does.

Original comment by: morgant

[jkbzh]

granthutchinson has provided me with further sample mbox files and we now have five that produce the bus error while loading the mbox, but four are successfully loaded & processed.

Original comment by: morgant