Re: Shared mappings not being inherited by children

[Agustina Arzille]

On 04/04/2016 04:34 AM, Samuel Thibault wrote:
> > > You can submit that patch to the libc-alpha@sourceware.org mailing list
> > > (explaining the rationale etc. alongside), putting roland@hack.frob.com
> > > and this bug-hurd@gnu.org list in Cc.  And we can commit the patch to
> > > the debian libc in parallel.
> > Done. Although, I had to use a different mail address for Roland, because
> > bugzilla wouldn't let me otherwise. Oh well, hope it isn't an issue.
> >
> > Anyway, here's the bug report for glibc:
> >
> > https://sourceware.org/bugzilla/show_bug.cgi?id=19903
> Mmm, in my memory entries in bugzilla don't get as much attention as
> patches submitted to libc-alpha.
>
> Please do as I said: send a complete patch to the libc-alpha mailing
> list, with roland in Cc, so he explictly gets the mail with the code
> inline. Doing things another way means more time from him, and thus much
> lower probability of getting an answer.

Allright, I did just that. Didn't know that bugzilla entries got less attention 
=/

> BTW, how did you discover this?  Does this trigger a failure in some
> application?

Not that I know of. I found out about this because I have a new libpthread
implementation based on the 'gsync' module, and I was testing process-shared
robust mutexes, and noticed they worked with shared memory as obtained
by 'shmget', but not 'mmap', so I dug around and found this bug.