Re: pkg/54348: pkgsrc net/fetch core dumps

[Hiroshi Hakoyama <hiroshi-hakoyama%nagano.ac.jp@localhost>]

Dear David,

Thank you for the reply.

> does the libfetch package match the fetch package?
Yes. I build fetch and libfetch after a new bootstrap.

> The important part here is probably not that it's Solaris but that
> it's sparc, and the code has probably done something either
> endian-dependent or that violates proper integer alignment.

Recently, I changed the IP address of the server from a global one to a local one.
I may break some configurations in the process.

Hiroshi





> 2019/07/08 5:25、David Holland <dholland-pbugs%netbsd.org@localhost>のメール:
> 
> The following reply was made to PR pkg/54348; it has been noted by GNATS.
> 
> From: David Holland <dholland-pbugs%netbsd.org@localhost>
> To: gnats-bugs%netbsd.org@localhost
> Cc: 
> Subject: Re: pkg/54348: pkgsrc net/fetch core dumps
> Date: Sun, 7 Jul 2019 20:24:42 +0000
> 
> On Sat, Jul 06, 2019 at 10:00:01AM +0000, hiroshi-hakoyama%nagano.ac.jp@localhost wrote:
>> /usr/pkg2/bin/fetch http://ftp.netbsd.org/robots.txt
>> Bus error (core dumped)
> 
> The important part here is probably not that it's Solaris but that
> it's sparc, and the code has probably done something either
> endian-dependent or that violates proper integer alignment.
> 
> That said, I can't see anything wrong in the code paths that match the
> stack trace, so I guess one question is: does the libfetch package
> match the fetch package? I could imagine something breaking if, say,
> the libfetch package was built years ago with a different compiler,
> but even that isn't all that likely.
> 
> It would be helpful if you (or someone who can reproduce the problem,
> perhaps on netbsd sparc) could recompile libfetch and fetch with
> debugging information and run them in the debugger to see exactly
> where the crash happens. :-|
> 
> -- 
> David A. Holland
> dholland%netbsd.org@localhost
>