----- Original Message -----
> From: Tatsuro MATSUOKA
> To: Achim Gratz; cygwin
> Cc:
> Date: 2019/3/6, Wed 10:10
> Subject: Re: Patch request to qt 5.9.4 (Re: [ANNOUNCEMENT] Qt 5.9.4)
>
>>> From: Achim Gratz
>
>>> To: cygwin
>>> Cc:
>>> Date: 2019/3/6, Wed 04:24
>>> Subject: Re: Patch request to qt 5.9.4 (Re: [ANNOUNCEMENT] Qt 5.9.4)
>>>
>>> T atsuro MATSUOKA writes:
>>>> I contacted with Enrico who tells me the patch that I have shown.
>>>
>>> Yes, but that monkeys around the real problem. What Yaakov is telling
>>> you is that there are two options and we don't know yet which one
> we
>>> have to deal with.
>>>
>>> If O_NONBLOCK is needed, but does not work correctly under Cygwin,
> then
>>> that's a bug in Cygwin. For this case, Corinna asks you to create
> an
>>> STC that demonstrates the bug so it can be triaged without having to
>>> work with something as complex as the Qt terminal in Cygwin.
>>>
>>> If on the other hand O_NONBLOCK is in fact not needed and just shows
>>> different behaviour on different systems without a bug in the
>>> implementation being present, then upstream should remove that
> argument.
>>>
>>>
>>> Regards,
>>> Achim.
>>
>> Seeing ML archive that Enrico indicated,
>>
>> Re: select() and named pipes
>>
>> #include <stdio.h>
>> #include <stdlib.h>
>> #include <errno.h>
>> #include <sys/select.h>
>> #include <sys/types.h>
>> #include <sys/stat.h>
>> #include <fcntl.h>
>> #include <unistd.h>
>>
>> #define FIFONAME "/tmp/pipe"
>>
>> int main(void)
>> {
>> int fd;
>> int nsel;
>> fd_set readfds;
>> FD_ZERO(&readfds);
>>
>> if (mkfifo(FIFONAME, 0600) < 0) {
>> perror("mkfifo");
>> exit(1);
>> }
>>
>> fd = open(FIFONAME, O_RDONLY | O_NONBLOCK);
>>
>> if (fd < 0) {
>> perror("open");
>> remove(FIFONAME);
>> exit(2);
>> }
>>
>> FD_SET(fd, &readfds);
>> do {
>> nsel = select(fd + 1, &readfds, 0, 0, 0);
>> } while (nsel == -1 && (errno == EINTR || errno == EAGAIN));
>>
>> if (nsel == -1) {
>> perror("select");
>> exit(3);
>> }
>>
>> if (FD_ISSET(fd, &readfds)) {
>> char buf[100];
>> int status;
>> int count = 0;
>> printf("%d: ", ++count);
>> while ((status = read(fd, buf, sizeof(buf) - 1))) {
>> if (status > 0) {
>> buf[status] = '\0';
>> printf("%s", buf);
>> } else if (errno == EAGAIN) {
>> printf("\n%d: ", ++count);
>> } else {
>> perror("read");
>> break;
>> }
>> }
>> }
>>
>> close(fd);
>> remove(FIFONAME);
>> return 0;
>> }
>>
>>
>>
>> As you wrote this is a issue of Cygwin issue.
>> But As long as O_NONBLOCK does not work on Cygwin, Enrico's workaround
> can
>> be attached for qt.
>
> Apart from Qt maintainer accepts Enrico's patch, I will raise O_NONBLOCK
> issue by another post.
I have updated the cygiwn by setup because cygwin is updated to 3.0.1-1.
The above code works with or without O_NONBLOCK.
On cygwin 3.0.0, with O_NONBLOCK, the code does not work.
Thus situation is changing.
But connection of gnuplot to qnuplot_qt still fails.
There seem to be Qt for Cygwin specific issue.
I cannot narrow down the issue at this moment.
I have to use Qt with debug symbols and gnuplot build with -g option and trace
the behaviors.
It take time and now I do not have enough time.
Please wait while.
Tatsuro
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
