Call for testing: OpenSSH 7.7

Jakub Jelen jjelen at redhat.com
Wed Mar 28 18:34:15 AEDT 2018


On Wed, 2018-03-28 at 10:54 +1100, Damien Miller wrote:
> On Tue, 27 Mar 2018, Jakub Jelen wrote:
> 
> > On Thu, 2018-03-22 at 14:42 +1100, Damien Miller wrote:
> > > Hi,
> > > 
> > > OpenSSH 7.7p1 is almost ready for release, so we would appreciate
> > > testing
> > > on as many platforms and systems as possible. This is a bugfix
> > > release.
> > > 
> > > [...]
> > > 
> > > Running the regression tests supplied with Portable OpenSSH does
> > > not
> > > require installation and is a simply:
> > > 
> > > $ ./configure && make tests
> > 
> > The localcommand test failed for me (for other reasons) and after
> > some
> > more investigation, I found that the following commit introduced an
> > pw
> > argument to the load_public_identity_files() function, but is
> > overwriting our local copy from main:
> > 
> > https://github.com/openssh/openssh-portable/commit/b7548b1
> > 
> > The following diff makes it using our copied pw variable instead of
> > overwriting it:
> 
> That should be removed, but I'm trying to understand how it could
> lead to problems. pw get overwritten locally, but AFAIK nothing
> from pw gets used after load_public_identity_files() exits.

I had applied a patch, that was freeing this memory once it was not
needed in the main() and noticed this issue while reviewing the code
what is going on with this variable and why it is "empty" in the
LocalCommand translations.

It showed up it is not the root cause, but worth mentioning.

Regards,
-- 
Jakub Jelen
Software Engineer
Security Technologies
Red Hat, Inc.


More information about the openssh-unix-dev mailing list