Jaromir Dolecek wrote:
> marius aamodt eriksen wrote:
> > in order to be able to preserve consistent semantics across poll,
> > select, and kqueue (EVFILT_READ), i propose the following change: on
> > EVFILT_READ, add an fflag NOTE_EOF which will return when the file
> > pointer *is* at the end of
> AFAIK yes for sockets, not for file descriptor (i.e. descriptor
> open to a file on filesystem). Would poll() give you read-availability
> event when on end of file on filesystem.
IIRC poll() is required to report that files are always readable.
(So tail -f can't use poll() to avoid looping.)
T
>> I think the difference is in the default behavior. When you're at
>> EOF, I know that poll() will give you a read-availability event, so
>> you'll read the EOF. Will kqueue?
> AFAIK yes for sockets, not for file descriptor (i.e. descriptor open
> to a file on filesystem). Would poll() give yo
Bill Studenmund wrote:
> I think the difference is in the default behavior. When you're at EOF, I
> know that poll() will give you a read-availability event, so you'll read
> the EOF. Will kqueue?
AFAIK yes for sockets, not for file descriptor (i.e. descriptor
open to a file on filesystem). Woul
marius aamodt eriksen wrote this message on Wed, Nov 12, 2003 at 14:06 -0500:
> > >right - the idea was to preserve existing semantics, while leaving the
> > >poll-like semantics optional.
> >
> > I'm not sure the "existing semantics" are anything more than a bug that
> > should be fixed.
>
> th
* Jason Thorpe <[EMAIL PROTECTED]> [031112 14:00]:
> On Nov 12, 2003, at 10:57 AM, marius aamodt eriksen wrote:
>
> >right - the idea was to preserve existing semantics, while leaving the
> >poll-like semantics optional.
>
> I'm not sure the "existing semantics" are anything more than a bug that
On Nov 12, 2003, at 10:57 AM, marius aamodt eriksen wrote:
right - the idea was to preserve existing semantics, while leaving the
poll-like semantics optional.
I'm not sure the "existing semantics" are anything more than a bug that
should be fixed.
-- Jason R. Thorpe <[EMAIL PROTECTED]>
* Jason Thorpe <[EMAIL PROTECTED]> [031112 13:54]:
> On Nov 12, 2003, at 10:40 AM, marius aamodt eriksen wrote:
>
> >correct - this is the difference, kqueue will not yield any event at
> >EOF.
>
> So, kqueue should simply be changed to report the event. I don't see
> any need for a separate E
On Nov 12, 2003, at 10:40 AM, marius aamodt eriksen wrote:
correct - this is the difference, kqueue will not yield any event at
EOF.
So, kqueue should simply be changed to report the event. I don't see
any need for a separate EOF flag. EOF can simply be determined as
normal in the kqueue case
* Bill Studenmund <[EMAIL PROTECTED]> [031112 12:27]:
> > I'm not sure I understand what is the exact issue.
>
> I'm only responding to the notes also.
>
> > Why would this be necessary or what does this exactly solve? AFAIK
> > poll() doesn't set any flags in this case neither, so I don't
> > s
On Wed, Nov 12, 2003 at 09:58:15AM +0100, Jaromir Dolecek wrote:
> marius aamodt eriksen wrote:
> > hi -
> >
> > in order to be able to preserve consistent semantics across poll,
> > select, and kqueue (EVFILT_READ), i propose the following change: on
> > EVFILT_READ, add an fflag NOTE_EOF which
marius aamodt eriksen wrote:
> hi -
>
> in order to be able to preserve consistent semantics across poll,
> select, and kqueue (EVFILT_READ), i propose the following change: on
> EVFILT_READ, add an fflag NOTE_EOF which will return when the file
> pointer *is* at the end of the file (effectively
12 matches
Mail list logo
