On Jan 21, 6:34pm, Corinna Vinschen wrote:
-- Subject: [Fwd: RE: ssh problem on Windows XP]
>
> is there any chance that we get a fix in the next couple of weeks?
I remain absolutely committed to fixing the problems that have been
reported, but I can't say that I'll have a fix in that timeframe,
I've been working on this, and I think I now understand the underlying
problem, but I don't yet have a fix.
The problem is that NtQueryInformationFile is stranger than I ever imagined.
First, some background, to be sure we're all speaking the same language:
Windows pipes actually seem more simila
On Sep 23, 2:31pm, Corinna Vinschen wrote:
-- Subject: Re: More pipe problems (was Re: [Fwd: 1.5.11-1: sftp performance
>
> It seems that NtQueryInformationFile doesn't return useful values
> anymore under XP SP2. I'm not quite sure though since it's the first
> time I'm looking into this issue.
On Sep 12, 4:42pm, Corinna Vinschen wrote:
-- Subject: Re: [Fwd: 1.5.11-1: sftp performance problem]
>
> Do I understand that right? sftp is in the blocking read on the pipe,
> there is data in the pipe and nevertheless read doesn't return? That's
> odd.
Yes, very. I'm still experimenting, so
> But native apps are not our first concern ...
OK, but native apps are a big concern for us (and perhaps others).
We use ...
ssh build-server-running-cygwin "make everything" > log-file
... and the Makefiles run lots of win32 native build tools.
> ... and the above can also easily be done
> I'll do some stracing to see what sftp is doing (that scp and ssh are not).
OK, I now have at least a partial understanding of what's happening, but
I need to do some additional experiments to learn more and work out a
solution.
sftp runs ssh, with a pair of pipes connecting the two programs,
o
On Sep 11, 3:02pm, Corinna Vinschen wrote:
-- Subject: Re: [Fwd: 1.5.11-1: sftp performance problem]
>
> http://cygwin.com/acronyms/#PCYMTNQREAIYR
Sorry ... done.
> | One really annoying consequence of socketpairs for sshd is that
> | Windows-native (i.e. non-Cygwin) programs don't know how to w
On Sep 10, 11:55am, [EMAIL PROTECTED] ("Bob Byrnes") wrote:
-- Subject: Re: [Fwd: 1.5.11-1: sftp performance problem]
>
> ... I don't use sftp much, but we pump *enormous* amounts
> of data through sshd otherwise, so it's odd that I haven'
On Sep 10, 11:01am, [EMAIL PROTECTED] (Corinna Vinschen) wrote:
-- Subject: [Fwd: 1.5.11-1: sftp performance problem]
>
> this problem is easily reproducible with sftp. It seems to be in some
> way related to your pipe patch.
>
> Due to speed considerations, Cygwin's implementation of OpenSSH use
On Sep 9, 10:06am, [EMAIL PROTECTED] (Christopher Faylor) wrote:
-- Subject: Re: [Patch] implementation of nonblocking writes on pipes
>
> Before we start adding more patches which are based on your previous work,
> could you reply to some of the problems raised in the cygwin mailing list?
Sure, I
cking writes on pipes are a bit complicated;
I haved tried to annotate the code with comments from the spec. If we
somehow inherit a tiny (non-Cygwin) pipe with size < PIPE_BUF, then we
try to do partial writes, as POSIX mandates for writes of size > PIPE_BUF.
--
Bob Byrnes
Curl Corporation
n-Cygwin) pipe, then we consider the
pipe writable only if it is completely empty, to minimize the probability
that a subsequent write will block.
--
Bob Byrnes
Curl Corporation
1 Cambridge Center, 10th Floor
Cambridge, MA 02142-1612
--------
2004-09-02 Bob Byrnes &
12 matches
Mail list logo
