mm/zswap: store <PAGE_SIZE compression failed page as-is
Subsystems (2)
PATCH v4 1/1 mm/zswap: store <PAGE_SIZE compression failed page as-is
When zswap writeback is enabled and it fails compressing a given page, the page is swapped out to the backing swap device. This behavior breaks the zswap's writeback LRU order, and hence users can experience unexpected latency spikes. If the page is compressed without failure, but results in a size of PAGE_SIZE, the LRU order is kept, but the decompression overhead for loading the page back on the later access is unnecessary.
Keep the LRU order and optimize unnecessary decompression overheads in those cases, by storing the original content as-is in zpool. The length field of zswap_entry will be set appropriately, as PAGE_SIZE. Hence whether it is saved as-is or not (whether decompression is unnecessary) is identified by 'zswap_entry->length == PAGE_SIZE'.
Because the uncompressed data is saved in zpool, same to the compressed ones, this introduces no change in terms of memory management including movability and migratability of involved pages.
This change is also not increasing per zswap entry metadata overhead. But as the number of incompressible pages increases, total zswap metadata overhead is proportionally increased. The overhead should not be problematic in usual cases, since the zswap metadata for single zswap entry is much smaller than PAGE_SIZE, and in common zswap use cases there should be a sufficient amount of compressible pages. Also it can be mitigated by the zswap writeback.
When the writeback is disabled, the additional overhead could be problematic. For the case, keep the current behavior that just returns the failure and let swap_writeout() put the page back to the active LRU list in the case.
Knowing how many compression failures from the crypto engine happened so far, and how many incompressible pages are stored at the given moment will be useful for future investigations. Add two new debugfs files, crypto_compress_fail and stored_incompressible_pages, for the two counts, respectively.
Tests
-----
I tested this patch using a simple self-written microbenchmark that is available at GitHub[1]. You can reproduce the test I did by executing run_tests.sh of the repo on your system. Note that the repo's documentation is not good as of this writing, so you may need to read and use the code.
The basic test scenario is simple. Run a test program making artificial accesses to memory having artificial content under memory.high-set memory limit and measure how many accesses were made in a given time.
The test program repeatedly and randomly access three anonymous memory regions. The regions are all 500 MiB size, and be accessed in the same probability. Two of those are filled up with a simple content that can easily be compressed, while the remaining one is filled up with a content that s read from /dev/urandom, which is easy to fail at compressing to a size smaller than PAGE_SIZE. The program runs for two minutes and prints out the number of accesses made every five seconds.
The test script runs the program under below four configurations.
- 0: memory.high is set to 2 GiB, zswap is disabled.
- 1-1: memory.high is set to 1350 MiB, zswap is disabled.
- 1-2: On 1-1, zswap is enabled without this patch.
- 1-3: On 1-2, this patch is applied.
For all zswap enabled cases, zswap shrinker is enabled.
Configuration '0' is for showing the original memory performance. Configurations 1-1, 1-2 and 1-3 are for showing the performance of swap, zswap, and this patch under a level of memory pressure (~10% of working set). Configurations 0 and 1-1 are not the main focus of this patch, but I'm adding those since their results transparently show how far this microbenchmark test is from the real world.
Because the per-5 seconds performance is not very reliable, I measured the average of that for the last one minute period of the test program run. I also measured a few vmstat counters including zswpin, zswpout, zswpwb, pswpin and pswpout during the test runs.
The measurement results are as below. To save space, I show performance numbers that are normalized to that of the configuration '0' (no memory pressure). The averaged accesses per 5 seconds of configuration '0' was 36493417.75.
config 0 1-1 1-2 1-3
perf_normalized 1.0000 0.0057 0.0235 0.0367
perf_stdev_ratio 0.0582 0.0652 0.0167 0.0346
zswpin 0 0 3548424 1999335
zswpout 0 0 3588817 2361689
zswpwb 0 0 10214 340270
pswpin 0 485806 772038 340967
pswpout 0 649543 144773 340270'perf_normalized' is the performance metric, normalized to that of configuration '0' (no pressure). 'perf_stdev_ratio' is the standard deviation of the averaged data points, as a ratio to the averaged metric value. For example, configuration '0' performance was showing 5.8% stdev. Configurations 1-1 and 1-3 were having about 6.5% and 6.1% stdev. Also the results were highly variable between multiple runs. So this result is not very stable but just showing ball park figures. Please keep this in your mind when reading these results.
Under about 10% of working set memory pressure, the performance was dropped to about 0.57% of no-pressure one, when the normal swap is used (1-1). Note that ~10% working set pressure is already extreme, at least on this test setup. No one would desire system setups that can degrade performance to 0.57% of the best case.
By turning zswap on (1-2), the performance was improved about 4x, resulting in about 2.35% of no-pressure one. Because of the incompressible pages in the third memory region, a significant amount of (non-zswap) swap I/O operations were made, though.
By applying this patch (1-3), about 56% performance improvement was made, resulting in about 3.67% of no-pressure one. Reduced pswpin of 1-3 compared to 1-2 let us see where this improvement came from.
Tests without Zswap Shrinker
----------------------------
Zswap shrinker is not enabled by default, so I ran the above test after disabling zswap shrinker. The results are as below.
config 0 1-1 1-2 1-3
perf_normalized 1.0000 0.0056 0.0185 0.0260
perf_stdev_ratio 0.0467 0.0348 0.1832 0.3387
zswpin 0 0 2506765 6049078
zswpout 0 0 2534357 6115426
zswpwb 0 0 0 0
pswpin 0 463694 472978 0
pswpout 0 686227 612149 0The overall normalized performance of the different configs are very similar to those of zswap shrinker enabled case. By adding the memory pressure, the performance was dropped to 0.56% of the original one. By enabling zswap without zswap shrinker, the performance was increased to 1.85% of the original one. By applying this patch on it, the performance was further increased to 2.6% of the original one.
Even though zswap shrinker is disabled, 1-2 shows high numbers of pswpin and pswpout because the incompressible pages are directly swapped out. In the case of 1-3, it shows zero pswpin and pswpout since it saves incompressible pages in the memory, and shows higher performance.
Note that the performance of 1-2 and 1-3 varies pretty much. Standard deviation of the performance for 1-2 was about 18.32% of the performance, while that for 1-3 was about 33.87%. Because zswap shrinker is disabled and the memory pressure is induced by memory.high, the workload got penalty_jiffies sleeps, and this resulted in the unstabilized performance.
Related Works
-------------
This is not an entirely new attempt. Nhat Pham and Takero Funaki tried very similar approaches in October 2023[2] and April 2024[3], respectively. The two approaches didn't get merged mainly due to the metadata overhead concern. I described why I think that shouldn't be a problem for this change, which is automatically disabled when writeback is disabled, at the beginning of this changelog.
This patch is not particularly different from those, and actually built upon those. I wrote this from scratch again, though. Hence adding Suggested-by tags for them. Actually Nhat first suggested this to me offlist.
Historically, writeback disabling was introduced partially as a way to solve the LRU order issue. Yosry pointed out[4] this is still suboptimal when the incompressible pages are cold, since the incompressible pages will continuously be tried to be zswapped out, and burn CPU cycles for compression attempts that will anyway fail. One imaginable solution for the problem is reusing the swapped-out page and its struct page to store in the zswap pool. But that's out of the scope of this patch.
[1] https://github.com/sjp38/eval_zswap/blob/master/run.sh
[2] https://lore.kernel.org/[email protected]
[3] https://lore.kernel.org/[email protected]
[4] https://lore.kernel.org/CAJD7tkZXS-UJVAFfvxJ0nNgTzWBiqepPYA4hEozi01_qktkitg@mail.gmail.com
---
Changes from v3
(https://lore.kernel.org/[email protected])
(discussions for changes from v3 were made on v2 thread)
- Drop the cumulated compression failure counter (compress_fail)
- Add a cumulated crypto-failure only counter (crypto_compress_fail)
- Add a not cumulated stored incompressible pages counter (stored_incompressible_pages)
- Cleanup compression failure handling code for readability
Changes from v2
(https://lore.kernel.org/[email protected])
- No code change bug changelog updates
- Add zswap shrinker disabled case test results.
- Fix a typo on changelog.
- Add a clarification of intention of 0 and 1-1 test configs.
Changes from v1
(https://lore.kernel.org/[email protected])
- Optimize out memcpy() per incompressible page saving, using k[un]map_local().
- Add a debugfs file for counting compression failures.
- Use a clear form of a ternary operation.
- Add the history of writeback disabling with a link.
- Wordsmith comments.
Changes from RFC v2
(https://lore.kernel.org/[email protected])
- Fix race conditions at decompressed pages identification.
- Remove the parameter and make saving as-is the default behavior.
- Open-code main changes.
- Clarify there is no memory management changes on the cover letter.
- Remove 20% pressure case from test results, since it is arguably too extreme and only adds confusion.
- Drop RFC tag.
Changes from RFC v1
(https://lore.kernel.org/[email protected])
- Consider PAGE_SIZE compression successes as failures.
- Use zpool for storing incompressible pages.
- Test with zswap shrinker enabled.
- Wordsmith changelog and comments.
- Add documentation of save_incompressible_pages parameter.
mm/zswap.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 54 insertions(+), 3 deletions(-)
mm/zswap.c +54 -3
@@ -42,8 +42,10 @@/********************************* * statistics **********************************/ -/* The number of compressed pages currently stored in zswap */ +/* The number of pages currently stored in zswap */ atomic_long_t zswap_stored_pages =ATOMIC_LONG_INIT (0 ); +/* The number of incompressible pages currently stored in zswap */ +atomic_long_t zswap_stored_incompressible_pages =ATOMIC_LONG_INIT (0 );/* * The statistics below are not protected from concurrent access for
@@ -60,6 +62,8 @@ static u64 zswap_written_back_pages;static u64 zswap_reject_reclaim_fail;/* Store failed due to compression algorithm failure */ static u64 zswap_reject_compress_fail; +/* Compression failed by the crypto library */ +static u64 zswap_crypto_compress_fail;/* Compressed page was too big for the allocator to (optimally) store */ static u64 zswap_reject_compress_poor;/* Load or writeback failed due to decompression failure */
@@ -811,6 +815,8 @@ static void zswap_entry_free(struct zswap_entry *entry)obj_cgroup_uncharge_zswap (entry->objcg, entry->length);obj_cgroup_put (entry->objcg); } +if (entry->length ==PAGE_SIZE ) +atomic_long_dec (&zswap_stored_incompressible_pages);zswap_entry_cache_free (entry);atomic_long_dec (&zswap_stored_pages); }
@@ -976,8 +982,28 @@ static bool zswap_compress(struct page *page, struct zswap_entry *entry,*/ comp_ret =crypto_wait_req (crypto_acomp_compress (acomp_ctx->req), &acomp_ctx->wait); dlen = acomp_ctx->req->dlen; -if (comp_ret) - goto unlock; + +/* +* If a page cannot be compressed into a size smaller than PAGE_SIZE, +* save the content as is without a compression, to keep the LRU order +* of writebacks. If writeback is disabled, reject the page since it +* only adds metadata overhead. swap_writeout() will put the page back +* to the active LRU list in the case. +*/ +if (comp_ret || !dlen) { + zswap_crypto_compress_fail++; + dlen =PAGE_SIZE ; + } +if (dlen >=PAGE_SIZE ) { +if (!mem_cgroup_zswap_writeback_enabled ( +folio_memcg (page_folio (page)))) { + comp_ret = -EINVAL ; + goto unlock; + } + comp_ret =0 ; + dlen =PAGE_SIZE ; + dst =kmap_local_page (page); + } zpool = pool->zpool; gfp =GFP_NOWAIT | __GFP_NORETRY | __GFP_HIGHMEM | __GFP_MOVABLE;
@@ -990,6 +1016,8 @@ static bool zswap_compress(struct page *page, struct zswap_entry *entry, entry->length = dlen; unlock: +if (dst != acomp_ctx->buffer) +kunmap_local (dst);if (comp_ret == -ENOSPC || alloc_ret == -ENOSPC ) zswap_reject_compress_poor++;else if (comp_ret)
@@ -1012,6 +1040,14 @@ static bool zswap_decompress(struct zswap_entry *entry, struct folio *folio) acomp_ctx =acomp_ctx_get_cpu_lock (entry->pool); obj =zpool_obj_read_begin (zpool, entry->handle, acomp_ctx->buffer); +/* zswap entries of length PAGE_SIZE are not compressed. */ +if (entry->length ==PAGE_SIZE ) { +memcpy_to_folio (folio,0 , obj, entry->length); +zpool_obj_read_end (zpool, entry->handle, obj); +acomp_ctx_put_unlock (acomp_ctx); +return true; + } +/* * zpool_obj_read_begin() might return a kmap address of highmem when * acomp_ctx->buffer is not used. However, sg_init_one() does not
@@ -1524,6 +1560,8 @@ static bool zswap_store_page(struct page *page,obj_cgroup_charge_zswap (objcg, entry->length); }atomic_long_inc (&zswap_stored_pages); +if (entry->length ==PAGE_SIZE ) +atomic_long_inc (&zswap_stored_incompressible_pages);/* * We finish initializing the entry while it's already in xarray.
@@ -1792,6 +1830,14 @@ static int debugfs_get_stored_pages(void *data, u64 *val) }DEFINE_DEBUGFS_ATTRIBUTE (stored_pages_fops, debugfs_get_stored_pages,NULL ,"%llu\n" ); +static int debugfs_get_stored_incompressible_pages (void *data,u64 *val) +{ + *val =atomic_long_read (&zswap_stored_incompressible_pages); +return 0 ; +} +DEFINE_DEBUGFS_ATTRIBUTE (stored_incompressible_pages_fops, + debugfs_get_stored_incompressible_pages,NULL ,"%llu\n" ); +static int zswap_debugfs_init (void ) {if (!debugfs_initialized ())
@@ -1809,6 +1855,8 @@ static int zswap_debugfs_init(void) zswap_debugfs_root, &zswap_reject_kmemcache_fail);debugfs_create_u64 ("reject_compress_fail" ,0444 , zswap_debugfs_root, &zswap_reject_compress_fail); +debugfs_create_u64 ("crypto_compress_fail" ,0444 , + zswap_debugfs_root, &zswap_crypto_compress_fail);debugfs_create_u64 ("reject_compress_poor" ,0444 , zswap_debugfs_root, &zswap_reject_compress_poor); debugfs_create_u64("decompress_fail", 0444,
@@ -1819,6 +1867,9 @@ static int zswap_debugfs_init(void) zswap_debugfs_root, NULL, &total_size_fops);debugfs_create_file ("stored_pages" ,0444 , zswap_debugfs_root,NULL , &stored_pages_fops); +debugfs_create_file ("stored_incompressible_pages" ,0444 , + zswap_debugfs_root,NULL , + &stored_incompressible_pages_fops);return 0 ; }
base-commit: 803d261a97f9b4025282723d2930e58d49adcbf9
Signature · 1 line
2.39.5
[...]
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 996
Quoted 12 lines
mm/zswap.c +12
-0+ +/* +* If a page cannot be compressed into a size smaller than PAGE_SIZE, +* save the content as is without a compression, to keep the LRU order +* of writebacks. If writeback is disabled, reject the page since it +* only adds metadata overhead. swap_writeout() will put the page back +* to the active LRU list in the case. +*/ +if (comp_ret || !dlen) { + zswap_crypto_compress_fail++; + dlen =PAGE_SIZE ; + }
I’m not entirely sure about this. As long as we pass 2*PAGE_SIZE as dst_buf, any error returned by crypto drivers should indicate a bug in the driver that needs to be fixed.
There have also been attempts to unify crypto drivers so they consistently return -ENOSPC when the destination buffer would overflow. If that has been achieved, we might be able to reduce the buffer from 2*PAGE_SIZE to just PAGE_SIZE in zswap. There was a long discussion on this here:
https://lore.kernel.org/linux-mm/[email protected]/
I’m not sure of the current status — do all crypto drivers now return a consistent -ENOSPC when the compressed size exceeds PAGE_SIZE? From what I recall during that discussion, most drivers already behaved this way, but Sergey Senozhatsky pointed out one or two exceptions.
Let’s sync with Herbert: have we reached the stage where all drivers reliably return -ENOSPC when dst_buf is PAGE_SIZE but the compressed size would exceed it?
Thanks
Barry
Quoted 4 lines · Barry Song
Let’s sync with Herbert: have we reached the stage where all drivers reliably return
-ENOSPCwhendst_bufisPAGE_SIZEbut the compressed size would exceed it?
It doesn't matter. Software compression should never fail, and if it does fail due to a bug, that's not something that the user of the API should worry about.
Hardware compression should always fall back to software compression in case of failure.
IOW all compression errors should be treated as incompressible data.
Cheers,
Signature · 3 lines
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Quoted 11 lines · Herbert Xu
Quoted 4 lines · Barry Song
Let’s sync with Herbert: have we reached the stage where all drivers reliably return
-ENOSPCwhendst_bufisPAGE_SIZEbut the compressed size would exceed it?It doesn't matter. Software compression should never fail, and if it does fail due to a bug, that's not something that the user of the API should worry about.
Quoted 5 lines
Hardware compression should always fall back to software compression in case of failure.
IOW all compression errors should be treated as incompressible data.
That is why I think it makes little sense to add a counter zswap_crypto_compress_fail, since zswap already passes 2*PAGE_SIZE as dst_buf, which is large enough to hold even incompressible data. Adding a counter that will always stay at zero seems meaningless.
We might want to revisit the old thread to check whether it is now safe for us to move to PAGE_SIZE in zswap now.
Thanks
Barry
Quoted 3 lines · Barry Song
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.
It's perfectly safe as LZO was fixed months ago.
Cheers,
Signature · 3 lines
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Nhat Pham
maintainerQuoted 7 lines · Herbert Xu
Quoted 3 lines · Barry Song
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.It's perfectly safe as LZO was fixed months ago.
Perfect. Then I'll revive Chengming's patch (see [1]) to reduce the compression buffer :)
[1]: https://lore.kernel.org/linux-mm/[email protected]/
Quoted 6 lines
Cheers,
--
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Quoted 5 lines · Nhat Pham
Quoted 3 lines · Herbert Xu
Quoted 1 line · Barry Song
Quoted 7 lines
Quoted 4 lines
Quoted 2 lines
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.It's perfectly safe as LZO was fixed months ago.
Perfect. Then I'll revive Chengming's patch (see [1]) to reduce the compression buffer :)
Nice!
But perhaps we should wait until SeongJae sends a new version that addresses the counter issue? Also, I noticed the following code may have problems with the patch:
if (comp_ret == -ENOSPC || alloc_ret == -ENOSPC)
zswap_reject_compress_poor++;Can we still reach the code comp_ret == -ENOSPC since we already handled comp_ret by ...
+ if (comp_ret || !dlen) {
+ zswap_crypto_compress_fail++;
+ dlen = PAGE_SIZE;
+ }
+ if (dlen >= PAGE_SIZE) {
+ if (!mem_cgroup_zswap_writeback_enabled(
+ folio_memcg(page_folio(page)))) {
+ comp_ret = -EINVAL;
+ goto unlock;
+ } + comp_ret = 0; + dlen = PAGE_SIZE; + dst = kmap_local_page(page);
+ }
Thanks
Barry
Quoted 11 lines · On Thu, 21 Aug 2025 18:27:52 +0800 Barry Song <[email protected]> wrote:
Quoted 10 lines · Nhat Pham
Quoted 7 lines · Herbert Xu
Quoted 3 lines · Barry Song
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.It's perfectly safe as LZO was fixed months ago.
Quoted 7 lines
Quoted 2 lines
Perfect. Then I'll revive Chengming's patch (see [1]) to reduce the compression buffer :)
Nice!
But perhaps we should wait until SeongJae sends a new version that addresses the counter issue?
Is there a reason to wait? I was thinking those are orthogonal problems?
Anyway, for the counter (crypto_compress_fail), I don't have a strong opinion. To my understanding, the options for path forward are...
1. remove it,
2. keep it as is, or
3. keep it, but account only -EINPROGRESS[1]
If I'm not missing other options, I'm tempted to the first option (remove it) since it doesn't change any existing things, and we can revisit later.
Please let me know if I'm missing another options or you have other preferrence on what option to take.
Quoted 13 lines
Also, I noticed the following code may
have problems with the patch:if (comp_ret == -ENOSPC || alloc_ret == -ENOSPC) zswap_reject_compress_poor++;Can we still reach the code
comp_ret==-ENOSPCsince we already handledcomp_retby ...(quoted diff) +4
-0+ if (comp_ret || !dlen) { + zswap_crypto_compress_fail++; + dlen = PAGE_SIZE; + }
Review
Quoted 10 lines
(quoted diff) +10
-0+ if (dlen >= PAGE_SIZE) { + if (!mem_cgroup_zswap_writeback_enabled( + folio_memcg(page_folio(page)))) { + comp_ret = -EINVAL; + goto unlock; + } + comp_ret = 0; + dlen = PAGE_SIZE; + dst = kmap_local_page(page); + }
Nice catch, thank you Berry. Actually there was a check for keeping the code reachable, but I forgot keeping it while updating the version from v3[2] to this one. I will make below change on the next version to restore it.
@@ -992,7 +992,7 @@ static bool zswap_compress(struct page *page, struct zswap_entry *entry,
if (dlen >= PAGE_SIZE) {
if (!mem_cgroup_zswap_writeback_enabled(
folio_memcg(page_folio(page)))) {
- comp_ret = -EINVAL;
+ comp_ret = comp_ret ? comp_ret : -EINVAL;
goto unlock;
}
comp_ret = 0;[1] https://lore.kernel.org/CAF8kJuPPsLzWu8+xm2A_UPHMBhb7OTjJNErM1Kp3hPmvHXNDUQ@mail.gmail.com
[2] https://lore.kernel.org/[email protected]
Thanks,
SJ
Quoted 27 lines · SeongJae Park
Quoted 11 lines · On Thu, 21 Aug 2025 18:27:52 +0800 Barry Song <[email protected]> wrote:
Quoted 10 lines · Nhat Pham
Quoted 7 lines · Herbert Xu
Quoted 3 lines · Barry Song
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.It's perfectly safe as LZO was fixed months ago.
Quoted 7 lines
Quoted 2 lines
Perfect. Then I'll revive Chengming's patch (see [1]) to reduce the compression buffer :)
Nice!
But perhaps we should wait until SeongJae sends a new version that addresses the counter issue?
Is there a reason to wait? I was thinking those are orthogonal problems?
Anyway, for the counter (
crypto_compress_fail), I don't have a strong opinion. To my understanding, the options for path forward are...
Quoted 6 lines
1. remove it,
2. keep it as is, or
3. keep it, but account only -EINPROGRESS[1]If I'm not missing other options, I'm tempted to the first option (remove it) since it doesn't change any existing things, and we can revisit later.
I am fine with 1) removing it. Maybe add a log once print error on the error code if -EINPROGRESS, just to know such extreme error has been triggered.
Quoted 2 lines
Please let me know if I'm missing other options or if you have other preferences.
I just don't want to hide the extreme error case but I am also fine with just removing it. It is your call.
Chris
Quoted 35 lines · On Thu, 21 Aug 2025 13:49:31 -0700 Chris Li <[email protected]> wrote:
Quoted 27 lines · SeongJae Park
Quoted 11 lines · On Thu, 21 Aug 2025 18:27:52 +0800 Barry Song <[email protected]> wrote:
Quoted 10 lines · Nhat Pham
Quoted 7 lines · Herbert Xu
Quoted 3 lines · Barry Song
We might want to revisit the old thread to check whether it is now safe for us to move to
PAGE_SIZEin zswap now.It's perfectly safe as LZO was fixed months ago.
Quoted 7 lines
Quoted 2 lines
Perfect. Then I'll revive Chengming's patch (see [1]) to reduce the compression buffer :)
Nice!
But perhaps we should wait until SeongJae sends a new version that addresses the counter issue?
Is there a reason to wait? I was thinking those are orthogonal problems?
Anyway, for the counter (
crypto_compress_fail), I don't have a strong opinion. To my understanding, the options for path forward are...Quoted 6 lines
1. remove it,
2. keep it as is, or
3. keep it, but account only -EINPROGRESS[1]If I'm not missing other options, I'm tempted to the first option (remove it) since it doesn't change any existing things, and we can revisit later.
Quoted 8 lines
I am fine with 1) removing it. Maybe add a log once print error on the error code if
-EINPROGRESS, just to know such extreme error has been triggered.Quoted 2 lines
Please let me know if I'm missing other options or if you have other preferences.
I just don't want to hide the extreme error case but I am also fine with just removing it. It is your call.
Thank you for your opinion, Chris! Unless others have different opinions, I will only remove the counter (option 1), since it is simplest and we can consider adding another counter or error logs on top of it.
Thanks,
SJ
[...]
Quoted 12 lines
Quoted 12 lines
Quoted 7 lines
1. remove it,
2. keep it as is, or
3. keep it, but account only -EINPROGRESS[1]If I'm not missing other options, I'm tempted to the first option (remove it) since it doesn't change any existing things, and we can revisit later.
I am fine with 1) removing it. Maybe add a log once print error on the error code if
-EINPROGRESS, just to know such extreme error has been triggered.Quoted 1 line
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Please let me know if I'm missing other options or if you have other preferences.
I just don't want to hide the extreme error case but I am also fine with just removing it. It is your call.
Thank you for your opinion, Chris! Unless others have different opinions, I will only remove the counter (option 1), since it is simplest and we can consider adding another counter or error logs on top of it.
Yes, that seems the best option—to remove the counter for now.
And I still need Herbert’s help to understand why crypto_wait_req() might return -EINPROGRESS, given the code below:
static inline int crypto_wait_req(int err, struct crypto_wait *wait)
{
switch (err) {
case -EINPROGRESS:
case -EBUSY:
wait_for_completion(&wait->completion);
reinit_completion(&wait->completion);
err = wait->err;
break;
}
return err;
}void crypto_req_done(void *data, int err)
{
struct crypto_wait *wait = data;
if (err == -EINPROGRESS)
return;
wait->err = err;
complete(&wait->completion);
}Is it even possible for crypto_wait_req() to return -EINPROGRESS, since crypto_req_done() will not call complete(&wait->completion) in that case at all?
Thanks
Barry
Quoted 4 lines · Barry Song
Is it even possible for
crypto_wait_req()to return-EINPROGRESS, sincecrypto_req_done()will not callcomplete(&wait->completion)in that case at all?
Of course crypto_wait_req cannot return -EINPROGRESS. However, I was responding to a request to make the crypto_acomp_compress call return a boolean instead of an error code. That is not possible because it has to be able to return -EINPROGRESS.
Nor is it possible to change crypto_wait_req to be boolean since other Crypto API operations (including decompression) can indeed fail.
Cheers,
Signature · 3 lines
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Quoted 7 lines · Herbert Xu
Quoted 4 lines · Barry Song
Is it even possible for
crypto_wait_req()to return-EINPROGRESS, sincecrypto_req_done()will not callcomplete(&wait->completion)in that case at all?
Quoted 9 lines
Of course
crypto_wait_reqcannot return-EINPROGRESS. However, I was responding to a request to make thecrypto_acomp_compresscall return a boolean instead of an error code. That is not possible because it has to be able to return-EINPROGRESS.Nor is it possible to change
crypto_wait_reqto be boolean since other Crypto API operations (including decompression) can indeed fail.
Ok. zswap only cares about the return value of crypto_wait_req(), not crypto_acomp_compress(). So it’s clear now—any error return means the page is incompressible.
Thanks
Barry
Quoted 14 lines · On Fri, 22 Aug 2025 12:48:27 +1200 Barry Song <[email protected]> wrote:
Quoted 14 lines
Quoted 14 lines
Quoted 7 lines
1. remove it,
2. keep it as is, or
3. keep it, but account only -EINPROGRESS[1]If I'm not missing other options, I'm tempted to the first option (remove it) since it doesn't change any existing things, and we can revisit later.
I am fine with 1) removing it. Maybe add a log once print error on the error code if
-EINPROGRESS, just to know such extreme error has been triggered.Quoted 2 lines
Please let me know if I'm missing other options or if you have other preferences.
Quoted 8 lines
Quoted 6 lines
Quoted 2 lines
I just don't want to hide the extreme error case but I am also fine with just removing it. It is your call.
Thank you for your opinion, Chris! Unless others have different opinions, I will only remove the counter (option 1), since it is simplest and we can consider adding another counter or error logs on top of it.
Yes, that seems the best option—to remove the counter for now.
Thank you for making your opinion clear, Barry. I will post v5 soon.
Thanks,
SJ
[...]
Quoted 16 lines · Barry Song
Quoted 14 lines · Herbert Xu
Quoted 4 lines · Barry Song
Let’s sync with Herbert: have we reached the stage where all drivers reliably return
-ENOSPCwhendst_bufisPAGE_SIZEbut the compressed size would exceed it?It doesn't matter. Software compression should never fail, and if it does fail due to a bug, that's not something that the user of the API should worry about.
Hardware compression should always fall back to software compression in case of failure.
Quoted 7 lines
Quoted 2 lines
IOW all compression errors should be treated as incompressible data.
That is why I think it makes little sense to add a counter
zswap_crypto_compress_fail, since zswap already passes 2*PAGE_SIZE asdst_buf, which is large enough to hold even incompressible data. Adding a counter that will always stay at zero seems meaningless.
I was thinking the crypto might be able to return errors other than the destination buffer being too small, here we give 2x the dst buffer already, shouldn't see a buffer to small error. In that case, having a counter mostly stay zero, only an exceptional error case is none zero is fine.
But if that error case can never happen, we should convert the crypto compression from returning error number to bool instead. It does not make sense to return a free form error code and also demand the caller does not check on it. If you insist that the caller should check return value like a boolean, just let the API return a boolean.
Chris
Quoted 6 lines · Chris Li
But if that error case can never happen, we should convert the crypto compression from returning error number to bool instead. It does not make sense to return a free form error code and also demand the caller does not check on it. If you insist that the caller should check return value like a boolean, just let the API return a boolean.
No we need the error value for other things, such as asynchronous return -EINPROGRESS.
Cheers,
Signature · 3 lines
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Quoted 11 lines · Herbert Xu
Quoted 6 lines · Chris Li
But if that error case can never happen, we should convert the crypto compression from returning error number to bool instead. It does not make sense to return a free form error code and also demand the caller does not check on it. If you insist that the caller should check return value like a boolean, just let the API return a boolean.
No we need the error value for other things, such as asynchronous return
-EINPROGRESS.
In that case it is fair game for the caller to check for the error other than -ENOSPC. If the return value is -EINPROGRESS during synchronous crypto compression, it is a bug. We are not encouraged to use WARN_ON, then having an error count to check that something truly unlikely but possible to happen is better than folding it into incompressible and pretend the error never happened. At least we can know such an error case happened.
Chris
Quoted 8 lines · Chris Li
In that case it is fair game for the caller to check for the error other than
-ENOSPC. If the return value is-EINPROGRESSduring synchronous crypto compression, it is a bug. We are not encouraged to useWARN_ON, then having an error count to check that something truly unlikely but possible to happen is better than folding it into incompressible and pretend the error never happened. At least we can know such an error case happened.
Sure warning on -EINPROGRESS is fine. But every other non-zero return value should be treated as incompressible.
Cheers,
Signature · 3 lines
Email: Herbert Xu <[email protected]> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
Quoted 11 lines · Herbert Xu
Quoted 8 lines · Chris Li
In that case it is fair game for the caller to check for the error other than
-ENOSPC. If the return value is-EINPROGRESSduring synchronous crypto compression, it is a bug. We are not encouraged to useWARN_ON, then having an error count to check that something truly unlikely but possible to happen is better than folding it into incompressible and pretend the error never happened. At least we can know such an error case happened.
Quoted 2 lines
Sure warning on
-EINPROGRESSis fine. But every other non-zero return value should be treated as incompressible.
Ack.
Thanks for the feedback.
Chris
Quoted 11 lines · Herbert Xu
Quoted 4 lines · Barry Song
Let’s sync with Herbert: have we reached the stage where all drivers reliably return
-ENOSPCwhendst_bufisPAGE_SIZEbut the compressed size would exceed it?It doesn't matter. Software compression should never fail, and if it does fail due to a bug, that's not something that the user of the API should worry about.
Quoted 5 lines
Hardware compression should always fall back to software compression in case of failure.
IOW all compression errors should be treated as incompressible data.
In that case, you seem to suggest that the crypto library can NEVER return an error other than the dst buffer is too small. In that case I would suggest the crypto library remove that free form error return code. Just return true for success compression or false on failure.
It is also strange to return a freeform error code and declare the caller shouldn't check on it. Ideally the caller shouldn't warn or bug on it, so it should handle the possible error other than -ENOSPC.
Chris
Quoted 15 lines · Barry Song
[...]
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 996
Quoted 12 lines
mm/zswap.c +12
-0+ +/* +* If a page cannot be compressed into a size smaller than PAGE_SIZE, +* save the content as is without a compression, to keep the LRU order +* of writebacks. If writeback is disabled, reject the page since it +* only adds metadata overhead. swap_writeout() will put the page back +* to the active LRU list in the case. +*/ +if (comp_ret || !dlen) { + zswap_crypto_compress_fail++; + dlen =PAGE_SIZE ; + }
Quoted 3 lines
I’m not entirely sure about this. As long as we pass 2*PAGE_SIZE as
dst_buf, any error returned by crypto drivers should indicate a bug in the driver that needs to be fixed.
That is what I have in mind for that counter, if that counter is not zero it is something with the crypto has gone wrong. If we are sure that it can never fail, maybe we shouldn't check the error return?
Quoted 6 lines
There have also been attempts to unify crypto drivers so they consistently return
-ENOSPCwhen the destination buffer would overflow. If that has been achieved, we might be able to reduce the buffer from 2*PAGE_SIZE to justPAGE_SIZEin zswap. There was a long discussion on this here:
https://lore.kernel.org/linux-mm/[email protected]/
Quoted 8 lines
I’m not sure of the current status — do all crypto drivers now return a consistent
-ENOSPCwhen the compressed size exceedsPAGE_SIZE? From what I recall during that discussion, most drivers already behaved this way, but Sergey Senozhatsky pointed out one or two exceptions.Let’s sync with Herbert: have we reached the stage where all drivers reliably return
-ENOSPCwhendst_bufisPAGE_SIZEbut the compressed size would exceed it?
I agree -ENOSPC should treat the compression ratio the same too low. However, is the crypto library able to return any error other than -ENOSPC? I am tempted to do something like BUG_ON() or WARN_ONCE() if other errors we think are never possible. However even the BUG_ON and WARN are discouraged in the kernel so we should handle the error. The one error is we can log it and monitor it to make sure it never happens.
If we are absolutely sure the other error should not happen. We should remove the free form error from the interface to reflect that. Make the function should just return bool success or failure as a buffer too small.
Chris
On Tue, 19 Aug 2025 12:34:04 -0700 SeongJae Park <[email protected]> wrote:
[...]
Quoted 5 lines
Knowing how many compression failures from the crypto engine happened so far, and how many incompressible pages are stored at the given moment will be useful for future investigations. Add two new debugfs files,
crypto_compress_failandstored_incompressible_pages, for the two counts, respectively.
[...]
mm/zswap.c line 48
Quoted 13 lines
mm/zswap.c +3
-1@@ -42,8 +42,10 @@/********************************* * statistics **********************************/ -/* The number of compressed pages currently stored in zswap */ +/* The number of pages currently stored in zswap */ atomic_long_t zswap_stored_pages =ATOMIC_LONG_INIT (0 ); +/* The number of incompressible pages currently stored in zswap */ +atomic_long_t zswap_stored_incompressible_pages =ATOMIC_LONG_INIT (0 );
Kernel test robot reported a sparse warning for the above line. Andrew, could you please add below attached fixup?
Thanks,
SJ
[...]
==== Attachment 0 (0001-mm-zswap-mark-zswap_stored_incompressible_pages-as-s.patch) ==== From 1d41d75c47a6d3ef3cbf58636faf2b4dc04616ba Mon Sep 17 00:00:00 2001
From: SeongJae Park <[email protected]>
Date: Thu, 21 Aug 2025 09:10:57 -0700
Subject: [PATCH] mm/zswap: mark zswap_stored_incompressible_pages as static
Only zswap.c uses zswap_stored_incompressible_pages, but it is not marked as static. This incurs a sparse warning that reported by kernel teset robot. Mark it as a static variable to eliminate the warning.
--- mm/zswap.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
mm/zswap.c +1 -1
@@ -45,7 +45,7 @@/* The number of pages currently stored in zswap */ atomic_long_t zswap_stored_pages =ATOMIC_LONG_INIT (0 );/* The number of incompressible pages currently stored in zswap */ -atomic_long_t zswap_stored_incompressible_pages =ATOMIC_LONG_INIT (0 ); +static atomic_long_t zswap_stored_incompressible_pages =ATOMIC_LONG_INIT (0 );/* * The statistics below are not protected from concurrent access for
Signature · 1 line
2.39.5
Quoted 277 lines · SeongJae Park
When zswap writeback is enabled and it fails compressing a given page, the page is swapped out to the backing swap device. This behavior breaks the zswap's writeback LRU order, and hence users can experience unexpected latency spikes. If the page is compressed without failure, but results in a size of
PAGE_SIZE, the LRU order is kept, but the decompression overhead for loading the page back on the later access is unnecessary.Keep the LRU order and optimize unnecessary decompression overheads in those cases, by storing the original content as-is in zpool. The length field of
zswap_entrywill be set appropriately, asPAGE_SIZE. Hence whether it is saved as-is or not (whether decompression is unnecessary) is identified by 'zswap_entry->length==PAGE_SIZE'.Because the uncompressed data is saved in zpool, same to the compressed ones, this introduces no change in terms of memory management including movability and migratability of involved pages.
This change is also not increasing per zswap entry metadata overhead. But as the number of incompressible pages increases, total zswap metadata overhead is proportionally increased. The overhead should not be problematic in usual cases, since the zswap metadata for single zswap entry is much smaller than
PAGE_SIZE, and in common zswap use cases there should be a sufficient amount of compressible pages. Also it can be mitigated by the zswap writeback.When the writeback is disabled, the additional overhead could be problematic. For the case, keep the current behavior that just returns the failure and let
swap_writeout()put the page back to the active LRU list in the case.Knowing how many compression failures from the crypto engine happened so far, and how many incompressible pages are stored at the given moment will be useful for future investigations. Add two new debugfs files,
crypto_compress_failandstored_incompressible_pages, for the two counts, respectively.Tests
-----I tested this patch using a simple self-written microbenchmark that is available at GitHub[1]. You can reproduce the test I did by executing
run_tests.shof the repo on your system. Note that the repo's documentation is not good as of this writing, so you may need to read and use the code.The basic test scenario is simple. Run a test program making artificial accesses to memory having artificial content under memory.high-set memory limit and measure how many accesses were made in a given time.
The test program repeatedly and randomly access three anonymous memory regions. The regions are all 500 MiB size, and be accessed in the same probability. Two of those are filled up with a simple content that can easily be compressed, while the remaining one is filled up with a content that s read from
/dev/urandom, which is easy to fail at compressing to a size smaller thanPAGE_SIZE. The program runs for two minutes and prints out the number of accesses made every five seconds.The test script runs the program under below four configurations.
- 0: memory.high is set to 2 GiB, zswap is disabled.
- 1-1: memory.high is set to 1350 MiB, zswap is disabled.
- 1-2: On 1-1, zswap is enabled without this patch.
- 1-3: On 1-2, this patch is applied.For all zswap enabled cases, zswap shrinker is enabled.
Configuration '0' is for showing the original memory performance. Configurations 1-1, 1-2 and 1-3 are for showing the performance of swap, zswap, and this patch under a level of memory pressure (~10% of working set). Configurations 0 and 1-1 are not the main focus of this patch, but I'm adding those since their results transparently show how far this microbenchmark test is from the real world.
Because the per-5 seconds performance is not very reliable, I measured the average of that for the last one minute period of the test program run. I also measured a few vmstat counters including zswpin, zswpout, zswpwb, pswpin and pswpout during the test runs.
The measurement results are as below. To save space, I show performance numbers that are normalized to that of the configuration '0' (no memory pressure). The averaged accesses per 5 seconds of configuration '0' was 36493417.75.
config 0 1-1 1-2 1-3 perf_normalized 1.0000 0.0057 0.0235 0.0367 perf_stdev_ratio 0.0582 0.0652 0.0167 0.0346 zswpin 0 0 3548424 1999335 zswpout 0 0 3588817 2361689 zswpwb 0 0 10214 340270 pswpin 0 485806 772038 340967 pswpout 0 649543 144773 340270'
perf_normalized' is the performance metric, normalized to that of configuration '0' (no pressure). 'perf_stdev_ratio' is the standard deviation of the averaged data points, as a ratio to the averaged metric value. For example, configuration '0' performance was showing 5.8% stdev. Configurations 1-1 and 1-3 were having about 6.5% and 6.1% stdev. Also the results were highly variable between multiple runs. So this result is not very stable but just showing ball park figures. Please keep this in your mind when reading these results.Under about 10% of working set memory pressure, the performance was dropped to about 0.57% of no-pressure one, when the normal swap is used (1-1). Note that ~10% working set pressure is already extreme, at least on this test setup. No one would desire system setups that can degrade performance to 0.57% of the best case.
By turning zswap on (1-2), the performance was improved about 4x, resulting in about 2.35% of no-pressure one. Because of the incompressible pages in the third memory region, a significant amount of (non-zswap) swap I/O operations were made, though.
By applying this patch (1-3), about 56% performance improvement was made, resulting in about 3.67% of no-pressure one. Reduced pswpin of 1-3 compared to 1-2 let us see where this improvement came from.
Tests without Zswap Shrinker
----------------------------Zswap shrinker is not enabled by default, so I ran the above test after disabling zswap shrinker. The results are as below.
config 0 1-1 1-2 1-3 perf_normalized 1.0000 0.0056 0.0185 0.0260 perf_stdev_ratio 0.0467 0.0348 0.1832 0.3387 zswpin 0 0 2506765 6049078 zswpout 0 0 2534357 6115426 zswpwb 0 0 0 0 pswpin 0 463694 472978 0 pswpout 0 686227 612149 0The overall normalized performance of the different configs are very similar to those of zswap shrinker enabled case. By adding the memory pressure, the performance was dropped to 0.56% of the original one. By enabling zswap without zswap shrinker, the performance was increased to 1.85% of the original one. By applying this patch on it, the performance was further increased to 2.6% of the original one.
Even though zswap shrinker is disabled, 1-2 shows high numbers of pswpin and pswpout because the incompressible pages are directly swapped out. In the case of 1-3, it shows zero pswpin and pswpout since it saves incompressible pages in the memory, and shows higher performance.
Note that the performance of 1-2 and 1-3 varies pretty much. Standard deviation of the performance for 1-2 was about 18.32% of the performance, while that for 1-3 was about 33.87%. Because zswap shrinker is disabled and the memory pressure is induced by memory.high, the workload got
penalty_jiffiessleeps, and this resulted in the unstabilized performance.Related Works
-------------This is not an entirely new attempt. Nhat Pham and Takero Funaki tried very similar approaches in October 2023[2] and April 2024[3], respectively. The two approaches didn't get merged mainly due to the metadata overhead concern. I described why I think that shouldn't be a problem for this change, which is automatically disabled when writeback is disabled, at the beginning of this changelog.
This patch is not particularly different from those, and actually built upon those. I wrote this from scratch again, though. Hence adding Suggested-by tags for them. Actually Nhat first suggested this to me offlist.
Historically, writeback disabling was introduced partially as a way to solve the LRU order issue. Yosry pointed out[4] this is still suboptimal when the incompressible pages are cold, since the incompressible pages will continuously be tried to be zswapped out, and burn CPU cycles for compression attempts that will anyway fail. One imaginable solution for the problem is reusing the swapped-out page and its struct page to store in the zswap pool. But that's out of the scope of this patch.
[1] https://github.com/sjp38/eval_zswap/blob/master/run.sh
[2] https://lore.kernel.org/[email protected]
[3] https://lore.kernel.org/[email protected]
[4] https://lore.kernel.org/CAJD7tkZXS-UJVAFfvxJ0nNgTzWBiqepPYA4hEozi01_qktkitg@mail.gmail.comSigned-off-by SeongJae Park <[email protected]>Suggested-by Nhat Pham <[email protected]>Suggested-by Takero Funaki <[email protected]>Acked-by Nhat Pham <[email protected]>Cc Chengming Zhou <[email protected]>Cc David Hildenbrand <[email protected]>Cc Johannes Weiner <[email protected]>Cc SeongJae Park <[email protected]>Cc Baoquan He <[email protected]>Cc Barry Song <[email protected]>Cc Chris Li <[email protected]>Cc Kairui Song <[email protected]>---
Changes from v3
(https://lore.kernel.org/[email protected])
(discussions for changes from v3 were made on v2 thread)
- Drop the cumulated compression failure counter (compress_fail)
- Add a cumulated crypto-failure only counter (crypto_compress_fail)
- Add a not cumulated stored incompressible pages counter (stored_incompressible_pages)
- Cleanup compression failure handling code for readabilityChanges from v2
(https://lore.kernel.org/[email protected])
- No code change bug changelog updates
- Add zswap shrinker disabled case test results.
- Fix a typo on changelog.
- Add a clarification of intention of 0 and 1-1 test configs.Changes from v1
(https://lore.kernel.org/[email protected])
- Optimize outmemcpy()per incompressible page saving, using k[un]map_local().
- Add a debugfs file for counting compression failures.
- Use a clear form of a ternary operation.
- Add the history of writeback disabling with a link.
- Wordsmith comments.Changes from RFC v2
(https://lore.kernel.org/[email protected])
- Fix race conditions at decompressed pages identification.
- Remove the parameter and make saving as-is the default behavior.
- Open-code main changes.
- Clarify there is no memory management changes on the cover letter.
- Remove 20% pressure case from test results, since it is arguably too extreme and only adds confusion.
- Drop RFC tag.Changes from RFC v1
(https://lore.kernel.org/[email protected])
- ConsiderPAGE_SIZEcompression successes as failures.
- Use zpool for storing incompressible pages.
- Test with zswap shrinker enabled.
- Wordsmith changelog and comments.
- Add documentation ofsave_incompressible_pagesparameter.mm/zswap.c | 57 +++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 54 insertions(+), 3 deletions(-)mm/zswap.c +7
-2@@ -42,8 +42,10 @@/********************************* * statistics **********************************/ -/* The number of compressed pages currently stored in zswap */ +/* The number of pages currently stored in zswap */ atomic_long_t zswap_stored_pages =ATOMIC_LONG_INIT (0 ); +/* The number of incompressible pages currently stored in zswap */ +atomic_long_t zswap_stored_incompressible_pages =ATOMIC_LONG_INIT (0 );/* * The statistics below are not protected from concurrent access for @@ -60,6 +62,8 @@ static u64 zswap_written_back_pages;static u64 zswap_reject_reclaim_fail;/* Store failed due to compression algorithm failure */ static u64 zswap_reject_compress_fail; +/* Compression failed by the crypto library */ +static u64 zswap_crypto_compress_fail;/* Compressed page was too big for the allocator to (optimally) store */ static u64 zswap_reject_compress_poor;/* Load or writeback failed due to decompression failure */ @@ -811,6 +815,8 @@ static void zswap_entry_free(struct zswap_entry *entry)obj_cgroup_uncharge_zswap (entry->objcg, entry->length);obj_cgroup_put (entry->objcg); } +if (entry->length ==PAGE_SIZE ) +atomic_long_dec (&zswap_stored_incompressible_pages);zswap_entry_cache_free (entry);atomic_long_dec (&zswap_stored_pages); }@@ -976,8 +982,28 @@ static bool zswap_compress(struct page *page, struct zswap_entry *entry,*/ comp_ret =crypto_wait_req (crypto_acomp_compress (acomp_ctx->req), &acomp_ctx->wait); dlen = acomp_ctx->req->dlen; -if (comp_ret)
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 993
Quoted 10 lines
mm/zswap.c +9
-1- goto unlock; + +/* +* If a page cannot be compressed into a size smaller than PAGE_SIZE, +* save the content as is without a compression, to keep the LRU order +* of writebacks. If writeback is disabled, reject the page since it +* only adds metadata overhead. swap_writeout() will put the page back +* to the active LRU list in the case. +*/ +if (comp_ret || !dlen) {
Looks good other than the feedback provided by Barry as well. Need to handle the -ENOSPC.
Other errors will depend on your plan to drop this counter or not. I will wait for your next version.
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 1001
Quoted 8 lines
mm/zswap.c +8
-0+ zswap_crypto_compress_fail++; + dlen =PAGE_SIZE ; + } +if (dlen >=PAGE_SIZE ) { +if (!mem_cgroup_zswap_writeback_enabled ( +folio_memcg (page_folio (page)))) { + comp_ret = -EINVAL ; + goto unlock;
I saw you mention this in the cover letter, so just to confirm we are on the same page. Current patch still has the issue [4] of write back disabled cases, the incompressible page will stay in the page LRU and possibly attempt to reclaim over and over again, right?
Chris
Quoted 80 lines
mm/zswap.c +30
-0+ } + comp_ret =0 ; + dlen =PAGE_SIZE ; + dst =kmap_local_page (page); + } zpool = pool->zpool; gfp =GFP_NOWAIT | __GFP_NORETRY | __GFP_HIGHMEM | __GFP_MOVABLE;@@ -990,6 +1016,8 @@ static bool zswap_compress(struct page *page, struct zswap_entry *entry, entry->length = dlen; unlock: +if (dst != acomp_ctx->buffer) +kunmap_local (dst);if (comp_ret == -ENOSPC || alloc_ret == -ENOSPC ) zswap_reject_compress_poor++;else if (comp_ret)@@ -1012,6 +1040,14 @@ static bool zswap_decompress(struct zswap_entry *entry, struct folio *folio) acomp_ctx =acomp_ctx_get_cpu_lock (entry->pool); obj =zpool_obj_read_begin (zpool, entry->handle, acomp_ctx->buffer); +/* zswap entries of length PAGE_SIZE are not compressed. */ +if (entry->length ==PAGE_SIZE ) { +memcpy_to_folio (folio,0 , obj, entry->length); +zpool_obj_read_end (zpool, entry->handle, obj); +acomp_ctx_put_unlock (acomp_ctx); +return true; + } +/* * zpool_obj_read_begin() might return a kmap address of highmem when * acomp_ctx->buffer is not used. However, sg_init_one() does not @@ -1524,6 +1560,8 @@ static bool zswap_store_page(struct page *page,obj_cgroup_charge_zswap (objcg, entry->length); }atomic_long_inc (&zswap_stored_pages); +if (entry->length ==PAGE_SIZE ) +atomic_long_inc (&zswap_stored_incompressible_pages);/* * We finish initializing the entry while it's already in xarray. @@ -1792,6 +1830,14 @@ static int debugfs_get_stored_pages(void *data, u64 *val) }DEFINE_DEBUGFS_ATTRIBUTE (stored_pages_fops, debugfs_get_stored_pages,NULL ,"%llu\n" ); +static int debugfs_get_stored_incompressible_pages (void *data,u64 *val) +{ + *val =atomic_long_read (&zswap_stored_incompressible_pages); +return 0 ; +} +DEFINE_DEBUGFS_ATTRIBUTE (stored_incompressible_pages_fops, + debugfs_get_stored_incompressible_pages,NULL ,"%llu\n" ); +static int zswap_debugfs_init (void ) {if (!debugfs_initialized ())@@ -1809,6 +1855,8 @@ static int zswap_debugfs_init(void) zswap_debugfs_root, &zswap_reject_kmemcache_fail);debugfs_create_u64 ("reject_compress_fail" ,0444 , zswap_debugfs_root, &zswap_reject_compress_fail); +debugfs_create_u64 ("crypto_compress_fail" ,0444 , + zswap_debugfs_root, &zswap_crypto_compress_fail);debugfs_create_u64 ("reject_compress_poor" ,0444 , zswap_debugfs_root, &zswap_reject_compress_poor); debugfs_create_u64("decompress_fail", 0444,@@ -1819,6 +1867,9 @@ static int zswap_debugfs_init(void) zswap_debugfs_root, NULL, &total_size_fops);debugfs_create_file ("stored_pages" ,0444 , zswap_debugfs_root,NULL , &stored_pages_fops); +debugfs_create_file ("stored_incompressible_pages" ,0444 , + zswap_debugfs_root,NULL , + &stored_incompressible_pages_fops);return 0 ; }base-commit:
803d261a97f9b4025282723d2930e58d49adcbf9--
2.39.5
Quoted 1 line · On Thu, 21 Aug 2025 14:21:11 -0700 Chris Li <[email protected]> wrote:
On Tue, Aug 19, 2025 at 12:34 PM SeongJae Park <[email protected]> wrote:
[...]
Quoted 12 lines
Quoted 12 lines
Historically, writeback disabling was introduced partially as a way to solve the LRU order issue. Yosry pointed out[4] this is still suboptimal when the incompressible pages are cold, since the incompressible pages will continuously be tried to be zswapped out, and burn CPU cycles for compression attempts that will anyway fail. One imaginable solution for the problem is reusing the swapped-out page and its struct page to store in the zswap pool. But that's out of the scope of this patch.
[1] https://github.com/sjp38/eval_zswap/blob/master/run.sh
[2] https://lore.kernel.org/[email protected]
[3] https://lore.kernel.org/[email protected]
[4] https://lore.kernel.org/CAJD7tkZXS-UJVAFfvxJ0nNgTzWBiqepPYA4hEozi01_qktkitg@mail.gmail.com
[...]
Quoted 9 lines
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 993
Quoted 8 lines
mm/zswap.c +8
-0+/* +* If a page cannot be compressed into a size smaller than PAGE_SIZE, +* save the content as is without a compression, to keep the LRU order +* of writebacks. If writeback is disabled, reject the page since it +* only adds metadata overhead. swap_writeout() will put the page back +* to the active LRU list in the case. +*/ +if (comp_ret || !dlen) {
Quoted 4 lines
Looks good other than the feedback provided by Barry as well. Need to handle the
-ENOSPC.
Other errors will depend on your plan to drop this counter or not. I will wait for your next version.
Ack. The next version will keep -ENOSPC comp_ret value so that reject_compress_poor counter is not broken, like I replied to Barry.
Quoted 10 lines
mm/zswap.c static bool zswap_compress(struct page *page, struct zswap_entry *entry, · line 1001
Quoted 10 lines
mm/zswap.c +8
-0+ zswap_crypto_compress_fail++; + dlen =PAGE_SIZE ; + } +if (dlen >=PAGE_SIZE ) { +if (!mem_cgroup_zswap_writeback_enabled ( +folio_memcg (page_folio (page)))) { + comp_ret = -EINVAL ; + goto unlock;
Quoted 4 lines
I saw you mention this in the cover letter, so just to confirm we are on the same page. Current patch still has the issue [4] of write back disabled cases, the incompressible page will stay in the page LRU and possibly attempt to reclaim over and over again, right?
You are correct. This patch is not making a change for writeback disabled cases.
Thanks,
SJ
[...]