From: Dick St.P. <stpeters@NetHeaven.com> - 2005-11-23 22:14:35
|
Murray S. Kucherawy writes: > On Tue, 22 Nov 2005, fre...@va... wrote: > > Is it a bug in sid-milter? > > Looks like a bug in libmarid. I'll see if I can find it. Don't look too hard. I think this may be an artifact of Fredrik having applied my 0.2.8 patch to 0.2.9. Unpatched 0.2.9 works ok for this case: in a fresh build directory without any of my patches applied except the utility patch: saint# ./sid-check -v 195.41.46.236 rei...@ma... sid-check for sid-milter version 0.2.9 Using asynchronous resolver ... sm_marid_check_host_frame: ip=<195.41.46.236>, domain=<mail.dk>, sender=<rei...@ma...> depth=[-1] Trying: <v=spf1 ptr:mail.dk ptr:tele.dk -all> Trying: <pfepb.post.tele.dk> Trying: <pfepb.post.tele.dk> spf=Pass ... However, in an older build directory that I think had my 0.2.8 patch applied: saint# ./sid-check.4 -v 195.41.46.236 rei...@ma... sid-check.4 for sid-milter version 0.2.9 sm_marid_check_host_frame: ip=<195.41.46.236>, domain=<mail.dk>, sender=<rei...@ma...> depth=[-1] Trying: <v=spf1 ptr:mail.dk ptr:tele.dk -all> spf=Fail (NotPermitted) ... Unfortunately I don't have any notes about how I built that .4 binary, so I have some work to do tracking this down. A 0.2.8-based sid-check with my 0.2.8 patch applied (and everything turned on) gives a pass. 0.2.8->0.2.9 changed only a tiny bit of code. However, that small change results in drastically altered callback sequences and program control flow. It was a high-leverage change. -- Dick St.Peters, stpeters@NetHeaven.com |