Changes in Kconfig

b261d222 - lib/crc: remove CONFIG_LIBCRC32C

Tue, 01 Apr 2025 22:16:00 +0000

lib/crc: remove CONFIG_LIBCRC32CNow that LIBCRC32C does nothing besides select CRC32, make every optionthat selects LIBCRC32C instead select CRC32 directly. Then removeLIBCRC32C.Reviewed-by: Christoph Hellwig Reviewed-by: "Martin K. Petersen" Acked-by: Ard Biesheuvel Link: https://lore.kernel.org/r/20250401221600.24878-8-ebiggers@kernel.orgSigned-off-by: Eric Biggers List of files: /linux-6.15/fs/erofs/Kconfig

0f24e3c0 - erofs: enable 48-bit layout support

Mon, 10 Mar 2025 09:56:25 +0000

erofs: enable 48-bit layout supportBoth 48-bit block addressing and encoded extents are implemented,let's enable them formally.Signed-off-by: Gao Xiang Acked-by: Chao Yu Link: https://lore.kernel.org/r/20250310095625.2623817-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

0d442ce0 - erofs: mark experimental fscache backend deprecated

Fri, 30 Aug 2024 03:28:40 +0000

erofs: mark experimental fscache backend deprecatedAlthough fscache is still described as "General Filesystem Caching" fornetwork filesystems and other things such as ISO9660 filesystems, it hasactually become a part of netfslib recently, which was unexpected at thetime when "EROFS over fscache" proposed (2021) since EROFS is entirely adisk filesystem and the dependency is redundant.Mark it deprecated and it will be removed after "fanotify pre-contenthooks" lands, which will provide the same functionality for EROFS.Reviewed-by: Sandeep Dhavale Reviewed-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20240830032840.3783206-4-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

fb176750 - erofs: add file-backed mount support

Fri, 30 Aug 2024 03:28:37 +0000

erofs: add file-backed mount supportIt actually has been around for years: For containers and other sandboxuse cases, there will be thousands (and even more) of authenticated(sub)images running on the same host, unlike OS images.Of course, all scenarios can use the same EROFS on-disk format, butbdev-backed mounts just work well for OS images since golden data isdumped into real block devices. However, it's somewhat hard forcontainer runtimes to manage and isolate so many unnecessary virtualblock devices safely and efficiently [1]: they just look like a burdento orchestrators and file-backed mounts are preferred indeed. Therewere already enough attempts such as Incremental FS, the originalComposeFS and PuzzleFS acting in the same way for immutable fses. Asfor current EROFS users, ComposeFS, containerd and Android APEXs willbe directly benefited from it.On the other hand, previous experimental feature "erofs over fscache"was once also intended to provide a similar solution (inspired byIncremental FS discussion [2]), but the following facts show file-backedmounts will be a better approach: - Fscache infrastructure has recently been moved into new Netfslib which is an unexpected dependency to EROFS really, although it originally claims "it could be used for caching other things such as ISO9660 filesystems too." [3] - It takes an unexpectedly long time to upstream Fscache/Cachefiles enhancements. For example, the failover feature took more than one year, and the deamonless feature is still far behind now; - Ongoing HSM "fanotify pre-content hooks" [4] together with this will perfectly supersede "erofs over fscache" in a simpler way since developers (mainly containerd folks) could leverage their existing caching mechanism entirely in userspace instead of strictly following the predefined in-kernel caching tree hierarchy.After "fanotify pre-content hooks" lands upstream to provide the samefunctionality, "erofs over fscache" will be removed then (as an EROFSinternal improvement and EROFS will not have to bother with on-demandfetching and/or caching improvements anymore.)[1] https://github.com/containers/storage/pull/2039[2] https://lore.kernel.org/r/CAOQ4uxjbVxnubaPjVaGYiSwoGDTdpWbB=w_AeM6YM=zVixsUfQ@mail.gmail.com[3] https://docs.kernel.org/filesystems/caching/fscache.html[4] https://lore.kernel.org/r/cover.1723670362.git.josef@toxicpanda.comCloses: https://github.com/containers/composefs/issues/144Reviewed-by: Sandeep Dhavale Reviewed-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20240830032840.3783206-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

7c35de4d - erofs: Zstandard compression support

Wed, 08 May 2024 23:44:53 +0000

erofs: Zstandard compression supportAdd Zstandard compression as the 4th supported algorithm since itbecomes more popular now and some end users have asked this forquite a while [1][2].Each EROFS physical cluster contains only one valid standardZstandard frame as described in [3] so that decompression can beperformed on a per-pcluster basis independently.Currently, it just leverages multi-call stream decompression APIs withinternal sliding window buffers. One-shot or bufferless decompressioncould be implemented later for even better performance if needed.[1] https://github.com/erofs/erofs-utils/issues/6[2] https://lore.kernel.org/r/Y08h+z6CZdnS1XBm@B-P7TQMD6M-0146.lan[3] https://www.rfc-editor.org/rfc/rfc8478.txtAcked-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20240508234453.17896-1-xiang@kernel.org List of files: /linux-6.15/fs/erofs/Kconfig

915cd30c - netfs, fscache: Combine fscache with netfs

Mon, 20 Nov 2023 15:55:18 +0000

netfs, fscache: Combine fscache with netfsNow that the fscache code is moved to be colocated with the netfslib codeso that they combined into one module, do the combining.Signed-off-by: David Howells Reviewed-by: Jeff Layton cc: Christian Brauner cc: linux-fsdevel@vger.kernel.orgcc: linux-cachefs@redhat.comcc: linux-nfs@vger.kernel.org,cc: linux-erofs@lists.ozlabs.org List of files: /linux-6.15/fs/erofs/Kconfig

62b241ef - MAINTAINERS: erofs: add EROFS webpage

Fri, 17 Nov 2023 08:53:29 +0000

MAINTAINERS: erofs: add EROFS webpageAdd a new `W:` field of the EROFS entry points to the documentationsite at .In addition, update the in-tree documentation and Kconfig too.Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20231117085329.1624223-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

798eecae - erofs: don't warn MicroLZMA format anymore

Sat, 21 Oct 2023 02:01:37 +0000

erofs: don't warn MicroLZMA format anymoreThe LZMA algorithm support has been landed for more than one year sinceLinux 5.16. Besides, the new XZ Utils 5.4 has been available in mostLinux distributions.Reviewed-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20231021020137.1646959-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

fd73a439 - erofs: boost negative xattr lookup with bloom filter

Sat, 22 Jul 2023 09:45:38 +0000

erofs: boost negative xattr lookup with bloom filterOptimise the negative xattr lookup with bloom filter.The bit value for the bloom filter map has a reverse semantics forcompatibility. That is, the bit value of 0 indicates existence, whilethe bit value of 1 indicates the absence of corresponding xattr.The initial version is _only_ enabled when xattr_filter_reserved iszero. The filter map internals may change in the future, in which casethe reserved flag will be set non-zero and we don't need bothering thecompatible bits again at that time. For now disable the optimization ifthis reserved flag is non-zero.Signed-off-by: Jingbo Xu Reviewed-by: Gao Xiang Reviewed-by: Chao Yu Link: https://lore.kernel.org/r/20230722094538.11754-3-jefflexu@linux.alibaba.comSigned-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

ffa09b3b - erofs: DEFLATE compression support

Thu, 10 Aug 2023 15:48:59 +0000

erofs: DEFLATE compression supportAdd DEFLATE compression as the 3rd supported algorithm.DEFLATE is a popular generic-purpose compression algorithm for quitelong time (many advanced formats like gzip, zlib, zip, png are allbased on that) as Apple documentation written "If you requireinteroperability with non-Apple devices, use COMPRESSION_ZLIB. [1]".Due to its popularity, there are several hardware on-market DEFLATEaccelerators, such as (s390) DFLTCC, (Intel) IAA/QAT, (HiSilicon) ZIPaccelerator, etc. In addition, there are also several high-performenceIP cores and even open-source FPGA approches available for DEFLATE.Therefore, it's useful to support DEFLATE compression in order to finda way to utilize these accelerators for asynchronous I/Os and getbenefits from these later.Besides, it's a good choice to trade off between compression ratiosand performance compared to LZ4 and LZMA. The DEFLATE core format issimple as well as easy to understand, therefore the code size of itsdecompressor is small even for the bootloader use cases. The runtimememory consumption is quite limited too (e.g. 32K + ~7K for each zlibstream). As usual, EROFS ourperforms similar approaches too.Alternatively, DEFLATE could still be used for some specific filessince EROFS supports multiple compression algorithms in one image.[1] https://developer.apple.com/documentation/compression/compression_algorithmReviewed-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20230810154859.118330-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

cf7f2732 - erofs: use HIPRI by default if per-cpu kthreads are enabled

Mon, 22 May 2023 09:21:41 +0000

erofs: use HIPRI by default if per-cpu kthreads are enabledAs Sandeep shown [1], high priority RT per-cpu kthreads aretypically helpful for Android scenarios to minimize the schedulinglatencies.Switch EROFS_FS_PCPU_KTHREAD_HIPRI on by default ifEROFS_FS_PCPU_KTHREAD is on since it's the typical use cases forEROFS_FS_PCPU_KTHREAD.Also clean up unneeded sched_set_normal().[1] https://lore.kernel.org/r/CAB=BE-SBtO6vcoyLNA9F-9VaN5R0t3o_Zn+FW8GbO6wyUqFneQ@mail.gmail.comReviewed-by: Yue Hu Reviewed-by: Sandeep Dhavale Reviewed-by: Chao Yu Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20230522092141.124290-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

3fffb589 - erofs: add per-cpu threads for decompression as an option

Wed, 08 Feb 2023 09:33:22 +0000

erofs: add per-cpu threads for decompression as an optionUsing per-cpu thread pool we can reduce the scheduling latency comparedto workqueue implementation. With this patch scheduling latency andvariation is reduced as per-cpu threads are high priority kthread_workers.The results were evaluated on arm64 Android devices running 5.10 kernel.The table below shows resulting improvements of total scheduling latencyfor the same app launch benchmark runs with 50 iterations. Schedulinglatency is the latency between when the task (workqueue kworker vskthread_worker) became eligible to run to when it actually startedrunning.+-------------------------+-----------+----------------+---------+| | workqueue | kthread_worker | diff |+-------------------------+-----------+----------------+---------+| Average (us) | 15253 | 2914 | -80.89% || Median (us) | 14001 | 2912 | -79.20% || Minimum (us) | 3117 | 1027 | -67.05% || Maximum (us) | 30170 | 3805 | -87.39% || Standard deviation (us) | 7166 | 359 | |+-------------------------+-----------+----------------+---------+Background: Boot times and cold app launch benchmarks are veryimportant to the Android ecosystem as they directly translate toresponsiveness from user point of view. While EROFS providesa lot of important features like space savings, we saw someperformance penalty in cold app launch benchmarks in few scenarios.Analysis showed that the significant variance was coming from thescheduling cost while decompression cost was more or less the same.Having per-cpu thread pool we can see from the above table that thisvariation is reduced by ~80% on average. This problem was discussedat LPC 2022. Link to LPC 2022 slides and talk at [1][1] https://lpc.events/event/16/contributions/1338/[ Gao Xiang: At least, we have to add this until WQ_UNBOUND workqueue issue [2] on many arm64 devices is resolved. ][2] https://lore.kernel.org/r/CAJkfWY490-m6wNubkxiTPsW59sfsQs37Wey279LmiRxKt7aQYg@mail.gmail.comSigned-off-by: Sandeep Dhavale Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20230208093322.75816-1-hsiangkao@linux.alibaba.com List of files: /linux-6.15/fs/erofs/Kconfig

c6be2bd0 - erofs: register fscache volume

Mon, 25 Apr 2022 12:21:33 +0000

erofs: register fscache volumeA new fscache based mode is going to be introduced for erofs, in whichcase on-demand read semantics is implemented through fscache.As the first step, register fscache volume for each erofs filesystem.That means, data blobs can not be shared among erofs filesystems. In thefollowing iteration, we are going to introduce the domain semantics, inwhich case several erofs filesystems can belong to one domain, and datablobs can be shared among these erofs filesystems of one domain.Signed-off-by: Jeffle Xu Reviewed-by: Gao Xiang Link: https://lore.kernel.org/r/20220425122143.56815-12-jefflexu@linux.alibaba.comAcked-by: Chao Yu Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

622ceadd - erofs: lzma compression support

Sun, 10 Oct 2021 21:31:45 +0000

erofs: lzma compression supportAdd MicroLZMA support in order to maximize compression ratios forspecific scenarios. For example, it's useful for low-end embeddedboards and as a secondary algorithm in a file for specific accesspatterns.MicroLZMA is a new container format for raw LZMA1, which was createdby Lasse Collin aiming to minimize old LZMA headers and get rid ofunnecessary EOPM (end of payload marker) as well as to enablefixed-sized output compression, especially for 4KiB pclusters.Similar to LZ4, inplace I/O approach is used to minimize runtimememory footprint when dealing with I/O. Overlapped decompression ishandled with 1) bounced buffer for data under processing or 2) extrashort-lived pages from the on-stack pagepool which will be shared inthe same read request (128KiB for example).Link: https://lore.kernel.org/r/20211010213145.17462-8-xiang@kernel.orgAcked-by: Chao Yu Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

dfeab2e9 - erofs: add multiple device support

Thu, 14 Oct 2021 08:10:10 +0000

erofs: add multiple device supportIn order to support multi-layer container images, add multipledevice feature to EROFS. Two ways are available to use for now: - Devices can be mapped into 32-bit global block address space; - Device ID can be specified with the chunk indexes format.Note that it assumes no extent would cross device boundary and mkfsshould take care of it seriously.In the future, a dedicated device manager could be introduced thenthus extra devices can be automatically scanned by UUID as well.Link: https://lore.kernel.org/r/20211014081010.43485-1-hsiangkao@linux.alibaba.comReviewed-by: Chao Yu Reviewed-by: Liu Bo Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

a08e67a0 - erofs: iomap support for non-tailpacking DIO

Thu, 05 Aug 2021 00:35:59 +0000

erofs: iomap support for non-tailpacking DIOAdd iomap support for non-tailpacking uncompressed data in order tosupport DIO and DAX.Direct I/O is useful in certain scenarios for uncompressed files.For example, double pagecache can be avoid by direct I/O whenloop device is used for uncompressed files containing upper layercompressed filesystem.This adds iomap DIO support for non-tailpacking cases first andtail-packing inline files are handled in the follow-up patch.Link: https://lore.kernel.org/r/20210805003601.183063-2-hsiangkao@linux.alibaba.comCc: linux-fsdevel@vger.kernel.orgReviewed-by: Chao Yu Signed-off-by: Huang Jianan Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

c5fcb511 - erofs: clean up file headers & footers

Wed, 02 Jun 2021 16:06:34 +0000

erofs: clean up file headers & footers - Remove my outdated misleading email address; - Get rid of all unnecessary trailing newline by accident.Link: https://lore.kernel.org/r/20210602160634.10757-1-xiang@kernel.orgReviewed-by: Chao Yu Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

9f6cc76e - erofs: introduce physical cluster slab pools

Wed, 07 Apr 2021 04:39:20 +0000

erofs: introduce physical cluster slab poolsSince multiple pcluster sizes could be used at once, the number ofcompressed pages will become a variable factor. It's necessary tointroduce slab pools rather than a single slab cache now.This limits the pclustersize to 1M (Z_EROFS_PCLUSTER_MAX_SIZE), andget rid of the obsolete EROFS_FS_CLUSTER_PAGE_LIMIT, which has nouse now.Link: https://lore.kernel.org/r/20210407043927.10623-4-xiang@kernel.orgAcked-by: Chao Yu Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

b858a484 - erofs: support superblock checksum

Mon, 04 Nov 2019 02:49:37 +0000

erofs: support superblock checksumIntroduce superblock checksum feature in order tocheck at mounting time.Note that the first 1024 bytes are ignore for x86boot sectors and other oddities.Link: https://lore.kernel.org/r/20191104024937.113939-1-gaoxiang25@huawei.comSigned-off-by: Pratik Shinde Reviewed-by: Chao Yu Cc: Dan Carpenter Signed-off-by: Gao Xiang List of files: /linux-6.15/fs/erofs/Kconfig

e2c71e74 - erofs: kill all erofs specific fault injection

Wed, 04 Sep 2019 02:09:06 +0000

erofs: kill all erofs specific fault injectionAs Christoph suggested [1], "Please just use plain kmalloceverywhere and let the normal kernel error injection codetake care of injeting any errors."[1] https://lore.kernel.org/r/20190829102426.GE20598@infradead.org/Reported-by: Christoph Hellwig Signed-off-by: Gao Xiang Link: https://lore.kernel.org/r/20190904020912.63925-20-gaoxiang25@huawei.comSigned-off-by: Greg Kroah-Hartman List of files: /linux-6.15/fs/erofs/Kconfig