On Fri, May 22, 2026 at 09:23:39AM +0100, Stuart Henderson wrote:
> From: Stuart Henderson <stu@spacehopper.org>
> To: renaud@openbsd.org
> Cc: ports@openbsd.org
> Subject: Re: [update] net/sniproxy 0.13.1
> Date: Fri, 22 May 2026 09:23:39 +0100
> 
> On 2026/05/22 08:49, renaud@openbsd.org wrote:
> > Hello,
> > 
> > Here is a very straightforward diff for net/sniproxy 0.13.1
> > Tested on amd64 and arm64
> > 
> > OK?
> 
> are the pledge hits in tests a problem? (203 is mlock)
> 
> make  check-TESTS
> PASS: address_test
> PASS: buffer_test
> PASS: cfg_tokenizer_test
> table_test[52026]: pledge "", syscall 203

Actually, no, it would only have caused an issue in the case
RAND_bytes() returned 0, which should never happen.
I have been running this in prod for 3 weeks without issue.

But you are right, it's better to correct that so here is 0.13.2


Index: Makefile
===================================================================
RCS file: /cvs/ports/net/sniproxy/Makefile,v
diff -u -p -r1.18 Makefile
--- Makefile	24 Mar 2026 18:21:42 -0000	1.18
+++ Makefile	22 May 2026 09:12:57 -0000
@@ -2,7 +2,7 @@ COMMENT =	name-based proxying of HTTPS w
 
 GH_ACCOUNT =	renaudallard
 GH_PROJECT =	sniproxy
-GH_TAGNAME =	0.11.1
+GH_TAGNAME =	0.13.2
 
 CATEGORIES =	net
 
Index: distinfo
===================================================================
RCS file: /cvs/ports/net/sniproxy/distinfo,v
diff -u -p -r1.13 distinfo
--- distinfo	24 Mar 2026 18:21:42 -0000	1.13
+++ distinfo	22 May 2026 09:12:57 -0000
@@ -1,2 +1,2 @@
-SHA256 (sniproxy-0.11.1.tar.gz) = Jr3T68Vmv40VKtwanNpjblcGgZAOGdmXAasdYPfnHfg=
-SIZE (sniproxy-0.11.1.tar.gz) = 426378
+SHA256 (sniproxy-0.13.2.tar.gz) = F1DmglXYFalDtPjG68bdAAbS7IOd+KIi6s3xITLFAD4=
+SIZE (sniproxy-0.13.2.tar.gz) = 439735