abort the current transmit attempt if our timeout is negative

- add patch from Filip Krska to abort a transmit attempt when we've given
  poll() a negative timeout (#838548)
This commit is contained in:
Nalin Dahyabhai 2012-09-10 16:30:11 -04:00
parent 4c51c8bc7e
commit a4ad97ae22
2 changed files with 21 additions and 1 deletions

View File

@ -0,0 +1,14 @@
Filip Krska's patch to abort the current pass after poll() returns when
we've given it an indefinite timeout.
--- krb5/src/lib/krb5/os/sendto_kdc.c
+++ krb5/src/lib/krb5/os/sendto_kdc.c
@@ -529,6 +529,8 @@ cm_select_or_poll(const struct select_st
*out = *in;
*sret = poll(out->fds, out->nfds, timeout);
e = SOCKET_ERRNO;
+ if (timeout < 0)
+ return e;
return (*sret < 0) ? e : 0;
#else
/* Use the select wrapper from cm.c. */

View File

@ -29,7 +29,7 @@
Summary: The Kerberos network authentication system
Name: krb5
Version: 1.10.3
Release: 4%{?dist}
Release: 5%{?dist}
# Maybe we should explode from the now-available-to-everybody tarball instead?
# http://web.mit.edu/kerberos/dist/krb5/1.10/krb5-1.10.3-signed.tar
Source0: krb5-%{version}.tar.gz
@ -85,6 +85,7 @@ Patch108: krb5-1.10.2-replay.patch
Patch109: krb5-1.10.3-as_rep_skew.patch
Patch110: krb5-1.10.3-keytab-etype-corners-prep.patch
Patch111: krb5-1.10.3-keytab-etype-corners.patch
Patch112: krb5-1.10.3-timeout_over.patch
License: MIT
URL: http://web.mit.edu/kerberos/www/
@ -282,6 +283,7 @@ ln -s NOTICE LICENSE
%patch109 -p1 -b .as_rep_skew
%patch110 -p1 -b .keytab-etype-corners-prep
%patch111 -p1 -b .keytab-etype-corners
%patch112 -p1 -b .timeout_over
rm src/lib/krb5/krb/deltat.c
gzip doc/*.ps
@ -851,6 +853,10 @@ exit 0
%changelog
* Fri Sep 6 2012 Nalin Dahyabhai <nalin@redhat.com> 1.10.3-5
- add patch from Filip Krska to abort a transmit attempt when we've given
poll() a negative timeout (#838548)
* Fri Sep 6 2012 Nalin Dahyabhai <nalin@redhat.com>
- on EL6, conflict with libsmbclient before 3.5.10-124, which is when it
stopped linking with a symbol which we no longer export (#771687)
- pull up patch for RT#7063, in which not noticing a prompt for a long