svn commit: trunk/uClibc/libc/sysdeps/linux/common

Mike Frysinger vapier at gentoo.org
Wed Jan 4 15:16:51 UTC 2006


On Wednesday 04 January 2006 09:57, Paul Mundt wrote:
> On Wed, Jan 04, 2006 at 09:50:23AM -0500, Mike Frysinger wrote:
> > On Wednesday 04 January 2006 04:06, Paul Mundt wrote:
> > > If the architecture doesn't provide _syscall6(), it's broken. mmap2()
> > > and depends on this, as do some others. If there's an architecture that
> > > can't implement this, there's going to have to be quite a few userspace
> > > hacks to accomodate it. grep for _syscall6() and pretty much everyone
> > > supports it anyways, if a new port doesn't support it, it's safe to
> > > assume the port is broken.
> >
> > i386 doesnt support it atm ;)
>
> In the kernel it most certainly does, whether someone bothered to update
> the uClibc files to reflect on that or not is another matter entirely. It
> was added during the 2.5 stage, so anything based off of a 2.4 port will
> be missing it.

right, i wasnt implying the limitation was anywhere other than uClibc
-mike



More information about the uClibc mailing list