Re: segfault in git-remote-http

From: Jeff King <peff@peff.net>
Date: Apr 10, 2013, 04:30 UTC
Message-ID: <20130410043030.GD795@sigill.intra.peff.net>
In-Reply-To: <20130409124044.908c160c718cc357a139354e@lavabit.com>

On Tue, Apr 09, 2013 at 12:40:44PM -0700, rh wrote:

Show 9 quoted lines

> > does not support hardlinks or symlinks). But I'm not sure which error
> > you are talking about. We can figure out inside the program which
> > program was invoked by checking argv, but I do not see us printing
> > remote-http anywhere.
> 
> I wasn't clear in my initial report....and may have omitted a
> significant fact. The "git clone" returned right away and I saw no
> error. The error shows up in dmesg via syslog, something like
> git-remote-http[1234]: segfault at xxxxxxxx blah blah in libcrypto

That message is not generated by git, but rather by the kernel, which pulls the name from argv[0], presumably. E.g., try:

  echo 'int main() { *(int *)0=0; }' >foo.c
  gcc foo.c
  ln a.out alternate
  ./a.out; ./alternate
  dmesg | tail -n 2

which should show both program names. Git invokes git-remote-* based on the URL you fed it. So if you are seeing a segfault in git-remote-http, presumably you fed it an http URL (which may still execute SSL code if it redirects to an https URL).

-Peff