On Wed, Jul 24, 2013 at 08:48:27AM +0200, Michal Simek wrote: > >> Create new CLONE_BACKWARDS3 type where stack_size is passed > >> via 3rd argument, parent thread id pointer via 4th, > >> child thread id pointer via 5th and tls value as 6th > >> argument > > > > I believe this also affects us in musl. What is the motivation for > > making a configure option that results in there being two incompatible > > syscall ABIs for the same arch? > > This sounds like a really bad idea... > > This patch fixes bug which was introduced by Al's patch where he moved > clone implementation from microblaze folder to generic location. > It means I am not creating two incompatible syscalls ABIs but fixing > broken one.
So this patch is just fixing a regression in the kernel? > > And how was glibc successfuly using a form that mismatched the > > existing kernel? Did nobody ever use/test it? > > We are running LTP syscall tests and there is not LTP test which > was able to find out this mismatch in clone. That's why I haven't > figure it out at that time and ACKed that origin patch. I would think pthread_create would have broken pretty badly; I remember early-on in porting musl to microblaze, we had the clone arguments misordered, and it blew up badly. ;) Perhaps you could just run some general libc/libpthread level tests to catch things like this that are hard to measure at the syscall-test level with existing tests. Rich -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/