On Jun 3 11:36, Warren Young wrote: > On 6/2/2013 04:31, Corinna Vinschen wrote: > >I dropped the F_MDLCK idea. Instead I implemented a specific fcntl code > >to switch to mandatory locking on a file descriptor: > > > > fcntl (fd, F_LCK_MANDATORY, 1); > > Could you add an O_MAND open(2) option as well to turn on the same > feature? This will avoid a race condition.
If you call F_LCK_MANDATORY right after open in the opening routine, there won't be (much of) a race. I know what you mean (it's the O_CLOEXEC dilemma) but is it really that important in terms of locking? > As I see it, the test will proceed in several stages: > > 0. I patch SQLite and run it against both .18 and your snapshot, and > strace it to verify that fcntl(fd, 0x99, 1) is called and returns 0 > with the snapshot and returns -1 with errno == EINVAL for .18. Both > count as "success" so we don't have to require .19. > > 1. I upload it, and Achim and Yaakov try it and see if this new > build satisfies their use cases. Well, we will have more than one round of testing, apparently...(*) Corinna (*) http://cygwin.com/ml/cygwin/2013-06/msg00042.html -- Corinna Vinschen Please, send mails regarding Cygwin to Cygwin Maintainer cygwin AT cygwin DOT com Red Hat -- 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