https://bugs.kde.org/show_bug.cgi?id=476887
--- Comment #7 from Paul Floyd <pjfl...@wanadoo.fr> --- Synchonising with aio_return shouldn't be too difficult. I'll do something along the same lines as Darwin but without the global 'was_a_successful_aio_read' variable and I'll put everything in the aio_return pre-syscall so there's no need to comminicate between aio_return pre- and post-syscall. Then I just need a testcase and it should be ready to go. I think I'll open a freebsd buzilla item for the manpages. The Linux one isn't fantastic but it's a lot clearer from the user's perspective. https://www.man7.org/linux/man-pages/man3/aio_read.3.html FreeBSD goes into a lot of implementation details https://man.freebsd.org/cgi/man.cgi?query=aio_read&apropos=0&sektion=2&manpath=FreeBSD+13.2-RELEASE+and+Ports&arch=default&format=html plus this is wrong "aio_read() will read iocb->aio_nbytes from the buffer pointed to by iocb->aio_buf" the read is from the fd into the buffer Last thing to do will be to take a look at the Valgrind doc. I think that 4.2.2 needs to say something about async reads https://valgrind.org/docs/manual/mc-manual.html#mc-manual.uninitvals -- You are receiving this mail because: You are watching all bug changes.
