Re: [PATCH 0/3] 64-bit futexes: Intro

[Ulrich Drepper <drepper@redhat.com>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Linus Torvalds wrote:
> We've had a 32-bit rwsemaphore in the kernel for a *long* time.

You misread what I wrote.

Semaphores are fine.  The problem are reader/writer locks where we need
three counters and some flags.  It is absolutely not acceptable to limit
the number of threads to 512 to make this fit.


> I think it's totally stupid for glibc to even *try* to do something like 
> this, since almost nobody will have a kernel with 64-bit futex support for 
> a long time in the wild anyway. So you need to support a 32-bit semaphore 
> in practice, and it's been done before.

Again, misread.  This is not functionality which is not available.
Semaphores are *of course* fine with 32 bits.  And there is a
reader/writer lock implementation.  It is just very slow compared to
what is possible.  The transition between systems miss the 64-bit
support and those which have it is completely transparent.  The same
glibc binary will work for both.

- --
➧ Ulrich Drepper ➧ Red Hat, Inc. ➧ 444 Castro St ➧ Mountain View, CA ❖
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkhAwpsACgkQ2ijCOnn/RHS7PACdGvdgI9pv+IcsSPkXPVwFJ5ZF
YNQAn01fm1wEc4EI/fwOVuM57XtCy9Dq
=LcpJ
-----END PGP SIGNATURE-----