> On Tue, Feb 02, 2021 at 01:12:04PM +0900, Changheun Lee wrote: > > > On Mon, Feb 01, 2021 at 11:52:48AM +0900, Changheun Lee wrote: > > > > > On Fri, Jan 29, 2021 at 12:49:08PM +0900, Changheun Lee wrote: > > > > > > bio size can grow up to 4GB when muli-page bvec is enabled. > > > > > > but sometimes it would lead to inefficient behaviors. > > > > > > in case of large chunk direct I/O, - 32MB chunk read in user space - > > > > > > all pages for 32MB would be merged to a bio structure if the pages > > > > > > physical addresses are contiguous. it makes some delay to submit > > > > > > until merge complete. bio max size should be limited to a proper size. > > > > > > > > > > > > When 32MB chunk read with direct I/O option is coming from userspace, > > > > > > kernel behavior is below now. it's timeline. > > > > > > > > > > > > | bio merge for 32MB. total 8,192 pages are merged. > > > > > > | total elapsed time is over 2ms. > > > > > > |------------------ ... ----------------------->| > > > > > > | 8,192 pages merged a bio. > > > > > > | at this time, first bio submit is done. > > > > > > | 1 bio is split to 32 read request and issue. > > > > > > |---------------> > > > > > > |---------------> > > > > > > |---------------> > > > > > > ...... > > > > > > |---------------> > > > > > > |--------------->| > > > > > > total 19ms elapsed to complete 32MB read done from device. | > > > > > > > > > > > > If bio max size is limited with 1MB, behavior is changed below. > > > > > > > > > bio_iov_iter_get_pages> > > | bio merge for 1MB. 256 pages are merged for each bio. > > > > > > | total 32 bio will be made. > > > > > > | total elapsed time is over 2ms. it's same. > > > > > > | but, first bio submit timing is fast. about 100us. > > > > > > |--->|--->|--->|---> ... -->|--->|--->|--->|--->| > > > > > > | 256 pages merged a bio. > > > > > > | at this time, first bio submit is done. > > > > > > | and 1 read request is issued for 1 bio. > > > > > > |---------------> > > > > > > |---------------> > > > > > > |---------------> > > > > > > ...... > > > > > > |---------------> > > > > > > |--------------->| > > > > > > total 17ms elapsed to complete 32MB read done from device. | > > > > > > > > > > Can you share us if enabling THP in your application can avoid this issue? BTW, you > > > > > need to make the 32MB buffer aligned with huge page size. IMO, THP perfectly fits > > > > > your case. > > > > > > > > > > > > > THP is enabled already like as below in my environment. It has no effect. > > > > > > > > cat /sys/kernel/mm/transparent_hugepage/enabled > > > > [always] madvise never > > > > > > The 32MB user buffer needs to be huge page size aligned. If your system > > > supports bcc/bpftrace, it is quite easy to check if the buffer is > > > aligned. > > > > > > > > > > > This issue was reported from performance benchmark application in open market. > > > > I can't control application's working in open market. > > > > It's not only my own case. This issue might be occured in many mobile environment. > > > > At least, I checked this problem in exynos, and qualcomm chipset. > > > > > > You just said it takes 2ms for building 32MB bio, but you never investigate the > > > reason. I guess it is from get_user_pages_fast(), but maybe others. Can you dig > > > further for the reason? Maybe it is one arm64 specific issue. > > > > > > BTW, bio_iov_iter_get_pages() just takes ~200us on one x86_64 VM with THP, which is > > > observed via bcc/funclatency when running the following workload: > > > > > > > I think you focused on bio_iov_iter_get_pages() because I just commented page > > merge delay only. Sorry about that. I missed details of this issue. > > Actually there are many operations during while-loop in do_direct_IO(). > > Page merge operation is just one among them. Page merge operation is called > > by dio_send_cur_page() in while-loop. Below is call stack. > > > > __bio_try_merge_page+0x4c/0x614 > > bio_add_page+0x40/0x12c > > dio_send_cur_page+0x13c/0x374 > > submit_page_section+0xb4/0x304 > > do_direct_IO+0x3d4/0x854 > > do_blockdev_direct_IO+0x488/0xa18 > > __blockdev_direct_IO+0x30/0x3c > > f2fs_direct_IO+0x6d0/0xb80 > > generic_file_read_iter+0x284/0x45c > > f2fs_file_read_iter+0x3c/0xac > > __vfs_read+0x19c/0x204 > > vfs_read+0xa4/0x144 > > > > 2ms delay is not only caused by page merge operation. it inculdes many the > > other operations too. But those many operations included page merge should > > be executed more if bio size is grow up. > > OK, got it. > > Then I think you can just limit bio size in dio_bio_add_page() instead of > doing it for all. Root cause is increasing of bio size after multipage bvec applying. So I think basic solution is limitation of bio size. Similar problem might be occurred in various scenario too. becasue bio merge function is called from many case in kernel. And bio max size was 1MB before applying of multipage bvec. So I think limitation of bio is reasonable. I think this solution is not bad for all. Becasue it just only activated by queue flag case by case. > > -- > Ming > > --- Changheun Lee Samsung Electronics
