OT - a good developers list?
All, would anyone care to recommend a good (active and knowledgeable) Linux developers mailing-list? I've got a situation where I am receiving EAGAIN from a blocking stream, something I did not expect to have to deal with. Or rather, if I have to be able to deal with it anyway, I might as well not use a blocking stream. Where does one go to post that type of questions? /Per Jessen, Zürich
On 3/1/06, Per Jessen
All,
would anyone care to recommend a good (active and knowledgeable) Linux developers mailing-list? I've got a situation where I am receiving EAGAIN from a blocking stream, something I did not expect to have to deal with. Or rather, if I have to be able to deal with it anyway, I might as well not use a blocking stream. Where does one go to post that type of questions?
Besides "suse-programming" there is "linux-c-programming" (majordomo@vger.kernel.org, http://vger.kernel.org/majordomo-info.html) specific to C programming with Linux (with no bias to any distribution) which is a highly recommended low-traffic list read by knowledgeable people. \Steve
Per Jessen wrote:
All,
would anyone care to recommend a good (active and knowledgeable) Linux developers mailing-list? I've got a situation where I am receiving EAGAIN from a blocking stream, something I did not expect to have to deal with. Or rather, if I have to be able to deal with it anyway, I might as well not use a blocking stream. Where does one go to post that type of questions?
Since what you're saying is impossible (blocking streams don't return EAGAIN) I would suggest making sure you haven't introduced O_NONBLOCK somewhere, and after that making an entry in a bugzilla somewhere
On 3/1/06, Anders Johansson
Per Jessen wrote:
All,
would anyone care to recommend a good (active and knowledgeable) Linux developers mailing-list? I've got a situation where I am receiving EAGAIN from a blocking stream, something I did not expect to have to deal with. Or rather, if I have to be able to deal with it anyway, I might as well not use a blocking stream. Where does one go to post that type of questions?
Since what you're saying is impossible (blocking streams don't return EAGAIN) I would suggest making sure you haven't introduced O_NONBLOCK somewhere, and after that making an entry in a bugzilla somewhere
I suppose, Per refers to a system call that supports blocking /and/ non-blocking operations on a stream but is not sure how it has been invoked. If O_NDELAY or O_NONBLOCK is set for such a stream it may cause a system call to return EAGAIN, but it's not blocking as you've pointed out correctly. \Steve
Steve Graegert wrote:
Since what you're saying is impossible (blocking streams don't return EAGAIN) I would suggest making sure you haven't introduced O_NONBLOCK somewhere, and after that making an entry in a bugzilla somewhere
I know it's _supposed_ to be impossible, but it's happening nonetheless. I even tried checking the flags of the stream just before trying to write to it, and although it is clearly blocking, it will occasionally respond with EAGAIN. Some googling led me to one or two postings referring to a similar problem.
I suppose, Per refers to a system call that supports blocking /and/ non-blocking operations on a stream but is not sure how it has been invoked.
It's a TCP socket on which I have opened a stream. For reading, it is nonblocking, for writing it's blocking. I use e.g. fputs() or fprintf() and will occasionally get a return of EOF with errno=EAGAIN. Oh, and of course I've actually got two identical file-descriptors, one for reading (nonblocking) and one for writing (blocking). /Per Jessen, Zürich
On 3/2/06, Per Jessen
Steve Graegert wrote:
Since what you're saying is impossible (blocking streams don't return EAGAIN) I would suggest making sure you haven't introduced O_NONBLOCK somewhere, and after that making an entry in a bugzilla somewhere
I know it's _supposed_ to be impossible, but it's happening nonetheless. I even tried checking the flags of the stream just before trying to write to it, and although it is clearly blocking, it will occasionally respond with EAGAIN. Some googling led me to one or two postings referring to a similar problem.
I suppose, Per refers to a system call that supports blocking /and/ non-blocking operations on a stream but is not sure how it has been invoked.
It's a TCP socket on which I have opened a stream. For reading, it is nonblocking, for writing it's blocking. I use e.g. fputs() or fprintf() and will occasionally get a return of EOF with errno=EAGAIN.
Oh, and of course I've actually got two identical file-descriptors, one for reading (nonblocking) and one for writing (blocking).
We could go to suse-programming and take a look at the code if you don't mind. \Steve
On 3/2/06, Per Jessen
Steve Graegert wrote:
We could go to suse-programming and take a look at the code if you don't mind.
suse-programming-e ? That's where I'm currently subscribed.
Great. Let us find a solution at this place. \Steve
participants (3)
-
Anders Johansson -
Per Jessen -
Steve Graegert