Hi Usama, > -----Original Message----- > From: Usama Arif <usamaarif642@xxxxxxxxx> > Sent: Monday, September 2, 2024 7:41 AM > To: Sridhar, Kanchana P <kanchana.p.sridhar@xxxxxxxxx>; linux- > kernel@xxxxxxxxxxxxxxx; linux-mm@xxxxxxxxx; hannes@xxxxxxxxxxx; > yosryahmed@xxxxxxxxxx; nphamcs@xxxxxxxxx; > chengming.zhou@xxxxxxxxx; ryan.roberts@xxxxxxx; Huang, Ying > <ying.huang@xxxxxxxxx>; 21cnbao@xxxxxxxxx; akpm@xxxxxxxxxxxxxxxxxxxx > Cc: Zou, Nanhai <nanhai.zou@xxxxxxxxx>; Feghali, Wajdi K > <wajdi.k.feghali@xxxxxxxxx>; Gopal, Vinodh <vinodh.gopal@xxxxxxxxx> > Subject: Re: [PATCH v6 0/3] mm: ZSWAP swap-out of mTHP folios > > > > On 29/08/2024 17:27, Kanchana P Sridhar wrote: > > Hi All, > > > > This patch-series enables zswap_store() to accept and store mTHP > > folios. The most significant contribution in this series is from the > > earlier RFC submitted by Ryan Roberts [1]. Ryan's original RFC has been > > migrated to v6.11-rc3 in patch 2/4 of this series. > > > > [1]: [RFC PATCH v1] mm: zswap: Store large folios without splitting > > https://lore.kernel.org/linux-mm/20231019110543.3284654-1- > ryan.roberts@xxxxxxx/T/#u > > > > Additionally, there is an attempt to modularize some of the functionality > > in zswap_store(), to make it more amenable to supporting any-order > > mTHPs. For instance, the function zswap_store_entry() stores a > zswap_entry > > in the xarray. Likewise, zswap_delete_stored_offsets() can be used to > > delete all offsets corresponding to a higher order folio stored in zswap. > > > > For accounting purposes, the patch-series adds per-order mTHP sysfs > > "zswpout" counters that get incremented upon successful zswap_store of > > an mTHP folio: > > > > /sys/kernel/mm/transparent_hugepage/hugepages-*kB/stats/zswpout > > > > A new config variable CONFIG_ZSWAP_STORE_THP_DEFAULT_ON (off by > default) > > will enable/disable zswap storing of (m)THP. When disabled, zswap will > > fallback to rejecting the mTHP folio, to be processed by the backing > > swap device. > > > > This patch-series is a precursor to ZSWAP compress batching of mTHP > > swap-out and decompress batching of swap-ins based on > swapin_readahead(), > > using Intel IAA hardware acceleration, which we would like to submit in > > subsequent RFC patch-series, with performance improvement data. > > > Hi Kanchana, > > If I am repeating any of the questions raised in previous revisions > over here, please feel free to just point to earlier responses! Sure, no problem. Thanks for the questions and observations with regards to the data posted in v6! > > Just wanted to check what does compress batching of mTHP swap-out > means? > Does it mean that zswap will not compress mTHP page by page, but will > compress the entire mTHP? > If it improves performance and possibly the numbers for case 2 below, maybe > its worth > adding it to this series? With Intel IAA, we have the opportunity to make use of compression and decompression engines in hardware to do parallel compressions during swapout and parallel decompressions during swapin with readahead. If compressions can be parallelized, we can improve reclaim performance. If decompressions can be parallelized, we can improve page-fault handling performance. We have implemented compress batching within mTHP folios during zswap store, as well as compress batching of any-order folios during shrink_folio_list() -- swap_writepage() using a plug mechanism, similar to the existing swap_write_unplug() implementation. Initially, our solution works at the granularity of compressing PAGE_SIZE pages within (many) folios in parallel, to maximize throughput with IAA and minimize latency per folio store/load. This is the compress/decompress batching I was referring to. To utilize IAA compress/decompress engines, we have developed the respective batching interfaces from shrink_folio_list() and from swapin_readahead(). Our experiments in multi-instance, highly contended scenarios under memory pressure, have demonstrated significant kernel and workload level performance improvements and overall system level memory savings. I was intending to submit this functionality as patch-series separate from the basic "mm: zswap: support mTHP swapout in zswap_store()" (this patch-series) as in my response to Yosry. As long as we can demonstrate that zswap-mTHP swapout is beneficial in and of itself, I believe we can submit IAA batching improvements as separate patch series, as noted in my response to Yosry. We are also staying tuned in to Barry Song's mTHP swapin efforts to eventually be able to swapout/swapin an mTHP as a single entity. In this case also, IAA byN can compress/decompress a tunable number of chunks of an mTHP in parallel [1]. The IAA byN approach is dependent on Barry's patchsets for mTHP swapin [2] and associated zsmalloc updates for storing larger compressed buffers [3]. Please note that Barry's work is focused on ZRAM/sync IO mTHP swapin and not for ZSWAP. [1] https://lore.kernel.org/all/8fe04e86f0907588d210885ac91965960f97f450.1714581792.git.andre.glover@xxxxxxxxxxxxxxx/T/#u [2] https://patchwork.kernel.org/project/linux-mm/cover/20240908232119.2157-1-21cnbao@xxxxxxxxx/ [3] https://lore.kernel.org/all/20240327214816.31191-1-21cnbao@xxxxxxxxx/ > > > Thanks to Ying Huang for pre-posting review feedback and suggestions! > > > > Thanks also to Nhat, Yosry and Barry for their helpful feedback, data > > reviews and suggestions! > > > > Changes since v5: > > ================= > > 1) Rebased to mm-unstable as of 8/29/2024, > > commit 9287e4adbc6ab8fa04d25eb82e097fed877a4642. > > 2) Added CONFIG_ZSWAP_STORE_THP_DEFAULT_ON (off by default) to > > enable/disable zswap_store() of mTHP folios. Thanks Nhat for the > > suggestion to add a knob by which users can enable/disable this > > change. Nhat, I hope this is along the lines of what you were > > thinking. > > 3) Added vm-scalability usemem data with 4K folios with > > CONFIG_ZSWAP_STORE_THP_DEFAULT_ON off, that I gathered to make > sure > > there is no regression with this change. > > 4) Added data with usemem with 64K and 2M THP for an alternate view of > > before/after, as suggested by Yosry, so we can understand the impact > > of when mTHPs are split into 4K folios in shrink_folio_list() > > (CONFIG_THP_SWAP off) vs. not split (CONFIG_THP_SWAP on) and stored > > in zswap. Thanks Yosry for this suggestion. > > > > Changes since v4: > > ================= > > 1) Published before/after data with zstd, as suggested by Nhat (Thanks > > Nhat for the data reviews!). > > 2) Rebased to mm-unstable from 8/27/2024, > > commit b659edec079c90012cf8d05624e312d1062b8b87. > > 3) Incorporated the change in memcontrol.h that defines obj_cgroup_get() if > > CONFIG_MEMCG is not defined, to resolve build errors reported by kernel > > robot; as per Nhat's and Michal's suggestion to not require a separate > > patch to fix the build errors (thanks both!). > > 4) Deleted all same-filled folio processing in zswap_store() of mTHP, as > > suggested by Yosry (Thanks Yosry!). > > 5) Squashed the commits that define new mthp zswpout stat counters, and > > invoke count_mthp_stat() after successful zswap_store()s; into a single > > commit. Thanks Yosry for this suggestion! > > > > Changes since v3: > > ================= > > 1) Rebased to mm-unstable commit > 8c0b4f7b65fd1ca7af01267f491e815a40d77444. > > Thanks to Barry for suggesting aligning with Ryan Roberts' latest > > changes to count_mthp_stat() so that it's always defined, even when THP > > is disabled. Barry, I have also made one other change in page_io.c > > where count_mthp_stat() is called by count_swpout_vm_event(). I would > > appreciate it if you can review this. Thanks! > > Hopefully this should resolve the kernel robot build errors. > > > > Changes since v2: > > ================= > > 1) Gathered usemem data using SSD as the backing swap device for zswap, > > as suggested by Ying Huang. Ying, I would appreciate it if you can > > review the latest data. Thanks! > > 2) Generated the base commit info in the patches to attempt to address > > the kernel test robot build errors. > > 3) No code changes to the individual patches themselves. > > > > Changes since RFC v1: > > ===================== > > > > 1) Use sysfs for zswpout mTHP stats, as per Barry Song's suggestion. > > Thanks Barry! > > 2) Addressed some of the code review comments that Nhat Pham provided > in > > Ryan's initial RFC [1]: > > - Added a comment about the cgroup zswap limit checks occuring once > per > > folio at the beginning of zswap_store(). > > Nhat, Ryan, please do let me know if the comments convey the summary > > from the RFC discussion. Thanks! > > - Posted data on running the cgroup suite's zswap kselftest. > > 3) Rebased to v6.11-rc3. > > 4) Gathered performance data with usemem and the rebased patch-series. > > > > > > Regression Testing: > > =================== > > I ran vm-scalability usemem 70 processes without mTHP, i.e., only 4K > > folios with mm-unstable and with this patch-series. The main goal was > > to make sure that there is no functional or performance regression > > wrt the earlier zswap behavior for 4K folios, > > CONFIG_ZSWAP_STORE_THP_DEFAULT_ON is not set, and zswap_store() of > 4K > > pages goes through the newly added code path [zswap_store(), > > zswap_store_page()]. > > > > The data indicates there is no regression. > > > > ------------------------------------------------------------------------------ > > mm-unstable 8-28-2024 zswap-mTHP v6 > > CONFIG_ZSWAP_STORE_THP_DEFAULT_ON > > is not set > > ------------------------------------------------------------------------------ > > ZSWAP compressor zstd deflate- zstd deflate- > > iaa iaa > > ------------------------------------------------------------------------------ > > Throughput (KB/s) 110,775 113,010 111,550 121,937 > > sys time (sec) 1,141.72 954.87 1,131.95 828.47 > > memcg_high 140,500 153,737 139,772 134,129 > > memcg_swap_high 0 0 0 0 > > memcg_swap_fail 0 0 0 0 > > pswpin 0 0 0 0 > > pswpout 0 0 0 0 > > zswpin 675 690 682 684 > > zswpout 9,552,298 10,603,271 9,566,392 9,267,213 > > thp_swpout 0 0 0 0 > > thp_swpout_ 0 0 0 0 > > fallback > > pgmajfault 3,453 3,468 3,841 3,487 > > ZSWPOUT-64kB-mTHP n/a n/a 0 0 > > SWPOUT-64kB-mTHP 0 0 0 0 > > ------------------------------------------------------------------------------ > > > > > > Performance Testing: > > ==================== > > Testing of this patch-series was done with the v6.11-rc3 mainline, without > > and with this patch-series, on an Intel Sapphire Rapids server, > > dual-socket 56 cores per socket, 4 IAA devices per socket. > > > > The system has 503 GiB RAM, with 176GiB ZRAM (35% of available RAM) as > the > > backing swap device for ZSWAP. zstd is configured as the ZRAM compressor. > > Core frequency was fixed at 2500MHz. > > > > The vm-scalability "usemem" test was run in a cgroup whose memory.high > > was fixed at 40G. The is no swap limit set for the cgroup. Following a > > similar methodology as in Ryan Roberts' "Swap-out mTHP without splitting" > > series [2], 70 usemem processes were run, each allocating and writing 1G of > > memory: > > > > usemem --init-time -w -O -n 70 1g > > > > The vm/sysfs mTHP stats included with the performance data provide > details > > on the swapout activity to ZSWAP/swap. > > > > Other kernel configuration parameters: > > > > ZSWAP Compressors : zstd, deflate-iaa > > ZSWAP Allocator : zsmalloc > > SWAP page-cluster : 2 > > > > In the experiments where "deflate-iaa" is used as the ZSWAP compressor, > > IAA "compression verification" is enabled. Hence each IAA compression > > will be decompressed internally by the "iaa_crypto" driver, the crc-s > > returned by the hardware will be compared and errors reported in case of > > mismatches. Thus "deflate-iaa" helps ensure better data integrity as > > compared to the software compressors. > > > > Throughput is derived by averaging the individual 70 processes' throughputs > > reported by usemem. sys time is measured with perf. All data points are > > averaged across 3 runs. > > > > Case 1: Baseline with CONFIG_THP_SWAP turned off, and mTHP is split in > reclaim. > > > ============================================================== > ================= > > > > In this scenario, the "before" is CONFIG_THP_SWAP set to off, that results in > > 64K/2M (m)THP to be split, and only 4K folios processed by zswap. > > > > The "after" is CONFIG_THP_SWAP set to on, and this patch-series, that > results > > in 64K/2M (m)THP to not be split, and processed by zswap. > > > > 64KB mTHP (cgroup memory.high set to 40G): > > ========================================== > > > > ------------------------------------------------------------------------------- > > v6.11-rc3 mainline zswap-mTHP Change wrt > > Baseline Baseline > > CONFIG_THP_SWAP=N CONFIG_THP_SWAP=Y > > ------------------------------------------------------------------------------- > > ZSWAP compressor zstd deflate- zstd deflate- zstd deflate- > > iaa iaa iaa > > ------------------------------------------------------------------------------- > > Throughput (KB/s) 136,113 140,044 140,363 151,938 3% 8% > > sys time (sec) 986.78 951.95 954.85 735.47 3% 23% > > memcg_high 124,183 127,513 138,651 133,884 > > memcg_swap_high 0 0 0 0 > > memcg_swap_fail 619,020 751,099 0 0 > > pswpin 0 0 0 0 > > pswpout 0 0 0 0 > > zswpin 656 569 624 639 > > zswpout 9,413,603 11,284,812 9,453,761 9,385,910 > > I would expect zswpout to either remain the same or slightly increase when > using > CONFIG_THP_SWAP. But for deflate-iaa, there is a 17% decrease in zswpout, > which > doesn't make sense? Good question. Without CONFIG_THP_SWAP, we see 751,099 memcg_swap_fail counts with deflate-iaa. With CONFIG_THP_SWAP, we see 0 memcg_swap_fail counts with deflate-iaa. My interpretation of this data is that with CONFIG_THP_SWAP, the main contributing factors to memcg.high breaches are faster swapout causing faster allocations + cgroup zswap charging. Without CONFIG_THP_SWAP, there seems to be an additional contribution of pages that remain in memory due to swap slot allocation failures; and hence more swapouts. Could there also be some effect of the reclaim path latency overhead of making 16 calls to swap_writepage() per mTHP that is split, vs. making one call in the case of zswap-mTHP? Would appreciate other analyses and explanations. > > > thp_swpout 0 0 0 0 > > thp_swpout_ 0 0 0 0 > > fallback > > pgmajfault 3,470 3,382 4,633 3,611 > > ZSWPOUT-64kB n/a n/a 590,768 586,521 > > SWPOUT-64kB 0 0 0 0 > > ------------------------------------------------------------------------------- > > > > > > 2MB PMD-THP/2048K mTHP (cgroup memory.high set to 40G): > > ======================================================= > > > > ------------------------------------------------------------------------------ > > v6.11-rc3 mainline zswap-mTHP Change wrt > > Baseline Baseline > > CONFIG_THP_SWAP=N CONFIG_THP_SWAP=Y > > ------------------------------------------------------------------------------ > > ZSWAP compressor zstd deflate- zstd deflate- zstd deflate- > > iaa iaa iaa > > ------------------------------------------------------------------------------ > > Throughput (KB/s) 164,220 172,523 165,005 174,536 0.5% 1% > > sys time (sec) 855.76 686.94 801.72 676.65 6% 1% > > memcg_high 14,628 16,247 14,951 16,096 > > memcg_swap_high 0 0 0 0 > > memcg_swap_fail 18,698 21,114 0 0 > > pswpin 0 0 0 0 > > pswpout 0 0 0 0 > > zswpin 663 665 5,333 781 > > zswpout 8,419,458 8,992,065 8,546,895 9,355,760 > > thp_swpout 0 0 0 0 > > thp_swpout_ 18,697 21,113 0 0 > > fallback > > pgmajfault 3,439 3,496 8,139 3,582 > > ZSWPOUT-2048kB n/a n/a 16,684 18,270 > > SWPOUT-2048kB 0 0 0 0 > > ----------------------------------------------------------------------------- > > > > We see improvements overall in throughput and sys time for zstd and > > deflate-iaa, when comparing before (THP_SWAP=N) vs. after > (THP_SWAP=Y). > > > > > > Case 2: Baseline with CONFIG_THP_SWAP enabled. > > ============================================== > > > > In this scenario, the "before" represents zswap rejecting mTHP, and the > mTHP > > being stored by the backing swap device. > > > > > Just curious, how did you make the before case of zswap rejecting mTHP > work? I suppose your question is about the experimental setup used for "before"? If so, the kernel I used was v6.11-rc3 in which zswap rejects mTHP stores, and mTHP gets processed in __swap_writepage(). For the v6 data, I had 176GiB ZRAM (35% of available RAM) as the backing swap device for ZSWAP. Hence the mTHPs would be processed by swap_writepage_bdev_sync(). Please let me know if this answers your question. > > > The "after" represents data with this patch-series, that results in 64K/2M > > (m)THP being processed by zswap. > > > > 64KB mTHP (cgroup memory.high set to 40G): > > ========================================== > > > > ------------------------------------------------------------------------------ > > v6.11-rc3 mainline zswap-mTHP Change wrt > > Baseline Baseline > > ------------------------------------------------------------------------------ > > ZSWAP compressor zstd deflate- zstd deflate- zstd deflate- > > iaa iaa iaa > > ------------------------------------------------------------------------------ > > Throughput (KB/s) 161,496 156,343 140,363 151,938 -13% -3% > > sys time (sec) 771.68 802.08 954.85 735.47 -24% 8% > > memcg_high 111,223 110,889 138,651 133,884 > > memcg_swap_high 0 0 0 0 > > memcg_swap_fail 0 0 0 0 > > pswpin 16 16 0 0 > > pswpout 7,471,472 7,527,963 0 0 > > zswpin 635 605 624 639 > > zswpout 1,509 1,478 9,453,761 9,385,910 > > thp_swpout 0 0 0 0 > > thp_swpout_ 0 0 0 0 > > fallback > > pgmajfault 3,616 3,430 4,633 3,611 > > ZSWPOUT-64kB n/a n/a 590,768 586,521 > > SWPOUT-64kB 466,967 470,498 0 0 > > ------------------------------------------------------------------------------ > > > > 2MB PMD-THP/2048K mTHP (cgroup memory.high set to 40G): > > ======================================================= > > > > ------------------------------------------------------------------------------ > > v6.11-rc3 mainline zswap-mTHP Change wrt > > Baseline Baseline > > ------------------------------------------------------------------------------ > > ZSWAP compressor zstd deflate- zstd deflate- zstd deflate- > > iaa iaa iaa > > ------------------------------------------------------------------------------ > > Throughput (KB/s) 192,164 194,643 165,005 174,536 -14% -10% > > sys time (sec) 823.55 830.42 801.72 676.65 3% 19% > > memcg_high 16,054 15,936 14,951 16,096 > > memcg_swap_high 0 0 0 0 > > memcg_swap_fail 0 0 0 0 > > pswpin 0 0 0 0 > > pswpout 8,629,248 8,628,907 0 0 > > zswpin 560 645 5,333 781 > > zswpout 1,416 1,503 8,546,895 9,355,760 > > thp_swpout 16,854 16,853 0 0 > > thp_swpout_ 0 0 0 0 > > fallback > > pgmajfault 3,341 3,574 8,139 3,582 > > ZSWPOUT-2048kB n/a n/a 16,684 18,270 > > SWPOUT-2048kB 16,854 16,853 0 0 > > ------------------------------------------------------------------------------ > > > > In the "Before" scenario, when zswap does not store mTHP, only allocations > > count towards the cgroup memory limit. However, in the "After" scenario, > > with the introduction of zswap_store() mTHP, both, allocations as well as > > the zswap compressed pool usage from all 70 processes are counted > towards > > the memory limit. As a result, we see higher swapout activity in the > > "After" data. Hence, more time is spent doing reclaim as the zswap cgroup > > charge leads to more frequent memory.high breaches. > > > > hmm, if that was the case, wouldn't "after" zswpout be much more than the > "before" > pswpout. But they look very similar? (Even goes down for zstd) For 64k mTHP, the "after" zswpout is considerably more than "before" pswpout, and so are the memcg_high counts. My comments were based on this (my apologies: I should have been more specific). In case of 2M THP, you are right: the "after" zswpout and "before" pswpout are quite similar. > > If pswpout in before is approximately equal to zswpout in after, then doesnt it > mean > that swap is performing better than zswap? which probably shouldnt happen. Agreed. Based on comments from Yosry and Nhat, I have posted 64k mTHP data with 4G SSD backing zswap, instead of 175G ZRAM backing zswap. If we agree to continue using 4G SSD as the backing device, I can gather data with 2M THP as well for further analysis of this patchset. Thanks, Kanchana > > Thanks, > Usama > > > This causes degradation in throughput and sys time with zswap mTHP, more > so > > in case of zstd than deflate-iaa. Compress latency could play a part in > > this - when there is more swapout activity happening, a slower compressor > > would cause allocations to stall for any/all of the 70 processes. > > > > In my opinion, even though the test set up does not provide an accurate > > way for a direct before/after comparison (because of zswap usage being > > counted in cgroup, hence towards the memory.high), it still seems > > reasonable for zswap_store to support (m)THP, so that further performance > > improvements can be implemented. > > > > One of the ideas that has shown promise in our experiments is to improve > > ZSWAP mTHP store performance using batching. With IAA > compress/decompress > > batching used in ZSWAP, we are able to demonstrate significant > > performance improvements and memory savings with IAA in scalability > > experiments, as compared to software compressors. We hope to submit > > this work as subsequent RFCs. > > > > I would greatly appreciate your code review comments and suggestions! > > > > Thanks, > > Kanchana > > > > [2] https://lore.kernel.org/linux-mm/20240408183946.2991168-1- > ryan.roberts@xxxxxxx/ > > > > > > Kanchana P Sridhar (3): > > mm: Define obj_cgroup_get() if CONFIG_MEMCG is not defined. > > mm: zswap: zswap_store() extended to handle mTHP folios. > > mm: swap: Count successful mTHP ZSWAP stores in sysfs mTHP zswpout > > stats. > > > > include/linux/huge_mm.h | 1 + > > include/linux/memcontrol.h | 4 + > > mm/Kconfig | 8 ++ > > mm/huge_memory.c | 3 + > > mm/page_io.c | 3 +- > > mm/zswap.c | 243 +++++++++++++++++++++++++++---------- > > 6 files changed, 200 insertions(+), 62 deletions(-) > > > > > > base-commit: 9287e4adbc6ab8fa04d25eb82e097fed877a4642
