On Wed, Jun 24, 2020 at 01:33:15AM -0700, Andrei Vagin wrote: > Allocate the time namespace page among VVAR pages and add the logic > to handle faults on VVAR properly. > > If a task belongs to a time namespace then the VVAR page which contains > the system wide VDSO data is replaced with a namespace specific page > which has the same layout as the VVAR page. That page has vdso_data->seq > set to 1 to enforce the slow path and vdso_data->clock_mode set to > VCLOCK_TIMENS to enforce the time namespace handling path. > > The extra check in the case that vdso_data->seq is odd, e.g. a concurrent > update of the VDSO data is in progress, is not really affecting regular > tasks which are not part of a time namespace as the task is spin waiting > for the update to finish and vdso_data->seq to become even again. > > If a time namespace task hits that code path, it invokes the corresponding > time getter function which retrieves the real VVAR page, reads host time > and then adds the offset for the requested clock which is stored in the > special VVAR page. >
> v2: Code cleanups suggested by Vincenzo. > v3: add a comment in __arch_get_timens_vdso_data. > v4: - fix an issue reported by the lkp robot. > - vvar has the same size with/without CONFIG_TIME_NAMESPACE, but the > timens page isn't allocated on !CONFIG_TIME_NAMESPACE. This > simplifies criu/vdso migration between different kernel configs. > v5: - Code cleanups suggested by Mark Rutland. > - In vdso_join_timens, mmap_write_lock is downgraded to > mmap_read_lock. The VMA list isn't changed there, zap_page_range > doesn't require mmap_write_lock. > > Reviewed-by: Vincenzo Frascino <vincenzo.frasc...@arm.com> > Reviewed-by: Dmitry Safonov <d...@arista.com> Hello Will and Catalin, Have you had a chance to look at this patch set? I think it is ready to be merged. Let me know if you have any questions. Thanks, Andrei
