fs/select.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/fs/select.c b/fs/select.c
index 08b91be..b0cf1f0 100644
--- a/fs/select.c
+++ b/fs/select.c
@@ -610,7 +610,7 @@ asmlinkage long sys_pselect7(int n, fd_set __user *inp, fd_set __user *outp,
sigprocmask(SIG_SETMASK, &ksigmask, &sigsaved);
}
- ret = core_sys_select(n, inp, outp, exp, &end_time);
+ ret = core_sys_select(n, inp, outp, exp, to);
ret = poll_select_copy_remaining(&end_time, tsp, 0, ret);
if (ret == -ERESTARTNOHAND) {
On Tue, Jan 13, 2009 at 9:29 AM, Bernd Schmidt <[email protected]> wrote:
> The patch below passes it to
> core_sys_select instead of &end_time, and this appears to fix the
> problem. Does this look like the correct approach?
Yes, that change is correct.
> Signed-off-by: Bernd Schmidt <[email protected]>
Reviewed-by: Ulrich Drepper <[email protected]>
On Tue 13 Jan 2009 14:38, Ulrich Drepper pondered:
> On Tue, Jan 13, 2009 at 9:29 AM, Bernd Schmidt <[email protected]> wrote:
> > The patch below passes it to
> > core_sys_select instead of &end_time, and this appears to fix the
> > problem. Does this look like the correct approach?
>
> Yes, that change is correct.
>
> > Signed-off-by: Bernd Schmidt <[email protected]>
>
> Reviewed-by: Ulrich Drepper <[email protected]>
Tested-by: Robin Getz <[email protected]>
I'm assuming this should go into stable then?