AgeCommit message (Collapse)Author
2014-12-04bcache-register: use register, not register_quietv1.0.8Gabriel
register_quiet doesn't report errors correctly, the following kernel patch needs to be applied: See for how this comes up in practice. Reverts 8327108eeaf3e0491b17d803da164c0827aae622
2014-12-04Replace bcache-register with c, use builtin kmodSimon Gomizelj
In a pure systemd pre-boot environment, register-bcache can't be shell code as there is no shell available. Switch to loading the module with udev's builtin kmod support and introduce a small binary for registering devices. Gabriel: Changed errors to be verbose
2014-05-10Do not compile bcache-test by defaultv1.0.7Gabriel
This way the Debian package can build without OpenSSL.
2014-05-10Take user-modified rules from /etcv1.0.6Nicholas Cull
Copy udev rules from /etc/udev/rules.d/ if they have been user-modified, otherwise copy from /lib/udev/rules.d/ instead. Follows usual udev conventions for modifying rules files.
2014-05-10udev: Skip floppies and cdromsGabriel
2014-05-10Correct typo in docsGabriel
2014-04-16Also accept flags with dashesGabriel
--data-offset didn't match the documentation. Follow gnu style and accept dashes everywhere.
2014-04-16Accept -o for --data-offsetKillian De Volder
This was mentioned in the help but had been omitted.
2014-01-26Update ignored filesGabriel
2014-01-26compile bcache-testAkira Hayakawa
Add -lm option to compile bcache-test. Signed-off-by: Akira Hayakawa <>
2014-01-26Fix typo "Spcifies"Vladimir Rutsky
2013-10-18Update the README to mention udev and initramfs hooksv1.0.5Gabriel
2013-10-18Add a mkinitcpio hookGabriel
Taken from the Arch Linux wiki, with some paths updated. Untested.
2013-10-16Install Dracut and initramfs-tools hooks unconditionallyGabriel
Distribution packagers may ignore tools that are not relevant.
2013-10-16udev: Skip DM devices marked privatev1.0.4Gabriel
2013-10-11Simplify 69-bcache.rulesv0.9Rolf Fokkens
Gabriel: simplified a bit further
2013-10-07Add an INSTALL macro to the Makefilev1.0.3Rolf Fokkens
2013-10-07Add DRACUTLIBDIR to MakefileRolf Fokkens
2013-10-07Add a dracut module for bcacheRolf Fokkens
2013-10-06Move probe-bcache to UDEVLIBDIRGabriel
2013-10-05udev: Fit into the standard rule sequencev1.0.2Gabriel
Move the rule to run a bit later, after the standard udev rules have called blkid. Don't run blkid manually, and run probe-bcache as a fallback which can be patched out in a package that depends on util-linux 2.24+. probe-bcache: bail if libblkid finds anything Preserve our last-resort safety check here in case earlier rules fail to call blkid.
2013-10-03super-show: Print labelGabriel
2013-09-27super-show: Show sync mode and cache replacement policyGabriel
2013-09-22Install the initramfs hook in /usr/share not /etcv1.0.1Gabriel
2013-09-22Use the absolute path to modprobeGabriel
This is necessary on Fedora, according to Rolf Fokkens.
2013-09-01make-bcache: Prevent stray superblocksGabriel
Fail if the device has any leftover superblock or partition table label. Add a --wipe-bcache flag to overwrite bcache superblocks and recommend wipefs for non-bcache superblocks. Once the device is safe to write, always zero immediately before the bcache superblock. Make sure errors go to stderr. Use pkg-config.
2013-08-26Use register_quiet for udev hookKent Overstreet
2013-07-21Don't register if a non-bcache superblock is found as wellGabriel
2013-07-17fix make-bcache to use logical_block_size, not stat.st_blocksizeLars Ellenberg
As discussed on irc yesterday. Cheers, Lars From 791a4b7b2c25e21ffeb4184da5e61f18cde86246 Mon Sep 17 00:00:00 2001 From: Lars Ellenberg <> Date: Wed, 17 Jul 2013 11:49:12 +0200 Subject: [PATCH] make-bcache: fix guessing of "block_size" to use BLKSSZGET Using stat.st_blocksize is wrong: that is the linux buffer cache granularity, it is tunable, and defaults to 4k. This caused bcache to default to create devices with logical_block_size = 4k, potentially breaking otherwise healthy disk or file system images unnecessarily. e.g. some xfs image previously created with 512 byte "sector size" would, once bcache'd, refuse to mount with mount: function not implemented device supports 4096 byte sectors (not 512) We want the logical_block_size here.
2013-07-17bcache-tools are GPLKent Overstreet
2013-06-09udev: Add /dev/bcache/by-label symlinks to cached devicesGabriel
2013-06-08udev: Persistent names for cached devicesGabriel
Add /dev/bcache/by-uuid/ symlinks to cached devices.
2013-06-07super-show: Minor manpage clarificationGabriel
2013-06-05super-show: Add a manpageGabriel
Taken from's debianisation.
2013-05-08super-show: show detached/clean/dirty/inconsistent statusGabriel
2013-05-08super-show: print cache mode (writethrough, writeback, writearound, none)Gabriel
2013-05-08super-show: don't show cache_countGabriel
This is a cache-set property, but not terribly useful. Showing it on backing devices was also incorrect.
2013-05-08super-show: show more cache-related informationGabriel
Only show nbuckets and nr_this_dev on cache devices; explain nbuckets' relation to device size by printing both total_sectors (superblock included) and cache_sectors (superblock excluded).
2013-05-08super-show: make the version annotation more terseGabriel
2013-04-30make-bdev: print the correct offset in the default caseGabriel
2013-04-23Get blocksize from the devices if it's not specifiedKent Overstreet
2013-04-15Update data offset format and warn about the previous one.Gabriel
Also update the macro to select a bdev. Also reindent.
2013-04-12use SB_JOURNAL_BUCKETS for superblockKent Overstreet
2013-04-12Trivial build fix.Gabriel
2013-04-11Add a flag to pick the cache set uuid.Gabriel
2013-04-11Register devices with just udev.Gabriel
This means bcache devices will be registered earlier and in all cases, not just when the rootfs fails to mount. The initramfs hook is still there to ensure the bcache module and udev rules are shipped if an initramfs is used.
2013-04-11Update for new superblock version for data_offsetKent Overstreet
2013-04-09Fix option parsing so that passing --bucket_size after the device worksKent Overstreet
This makes supporting -U more annoying; dropping that for the moment, unless someone complains
2013-03-12Use uppercase hex when showing csum and expected.Gabriel
2013-03-10Show the writeback flag and the discard flag.Gabriel