Re: Bug#71503: apt: [hurd] doesn't support MS_SYNC
On Tue, Sep 12, 2000 at 06:52:47PM -0600, Jason Gunthorpe wrote:
>
> On Wed, 13 Sep 2000 Marcus.Brinkmann@ruhr-uni-bochum.de wrote:
>
> > the Hurd doesn't support MS_SYNC on mmaped memory objects, sorry,
> > and it is not easy to add (because of gnumach shortage). Please add the
> > following workaround: Replace the two occurences of
>
> What!? I hope you have a copy of the POSIX spec to back up your claim
> that it is valid to define _POSIX_SYNC_IO and not implement the whole SYNC
> IO interface.
Please take it slowly, there is no reason for a heart attack.
I never claimed that the Hurd is doing the right thing here.
* #47896: MS_SYNC Macros Missing from Hurd development environment
Package: libc0.2-dev; Reported by: Brent Fulgham <brent.fulgham@xpsystems.com>;
328 days old
> AFAIK your glibc is broken, if it doesn't implement the msync syscall
> as defined by POSIX it must not define _POSIX_SYNC_IO.
This possibility occured to me, and because I don't have a copy of the POSIX
standard to check, I mailed to the bug-hurd mailing list in parallel with the
bug report to clarify if we are doing bad here:
>From http://www.geocrawler.com/lists/3/GNU/332/0/4333044/
> Hi,
> is it posixly correct to define _POSIX_SYNCHRONIZED_IO although we don't
> support msync?
If you want to put this report on hold until we get an answer on that
(most likely from Roland McGrath), that's fine with me. I will keep you up
to date.
Thanks,
Marcus
--
`Rhubarb is no Egyptian god.' Debian http://www.debian.org Check Key server
Marcus Brinkmann GNU http://www.gnu.org for public PGP Key
Marcus.Brinkmann@ruhr-uni-bochum.de, marcus@gnu.org PGP Key ID 36E7CD09
http://homepage.ruhr-uni-bochum.de/Marcus.Brinkmann/ brinkmd@debian.org
Reply to: