summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* cmake: Fixed wrong package name in the Fuse find moduleManuel Bentele2020-08-191-2/+2
|
* [SERVER] Fix rid 0 proxy lookup if local version is newerSimon Rettberg2020-08-141-5/+15
| | | | | There was a logic bug that would favor a remotely looked up image rid, even if we already found a higher revision locally.
* [SERVER] FUSE: Set name for fuse theadsSimon Rettberg2020-07-311-0/+2
|
* [SERVER] FUSE: Use splice with move for local readsSimon Rettberg2020-07-311-2/+16
| | | | | | Greatly improves performance when reading from the local disk cache, and additionally avoids having everything in the fs cache twice (disk view and fuse view).
* [SERVER] FUSE: Fix arguments to fuseSimon Rettberg2020-07-302-2/+3
|
* [SERVER] Remove dead codeSimon Rettberg2020-07-301-6/+0Star
|
* [SERVER] Fix: Missing argument to fprintf()Simon Rettberg2020-07-301-1/+1
|
* [SERVER] Inline imageLookup/Open since we use fuse_mtSimon Rettberg2020-07-301-56/+34Star
|
* [SERVER] FUSE: Add readdir for /Simon Rettberg2020-07-301-1/+40
|
* [SERVER] Add FUSE modeSimon Rettberg2020-07-2812-103/+795
| | | | | Still needs some cleanup and optimizations, variable naming sucks, comments, etc.
* Merge branch 'no-working-flag' into fuse_llSimon Rettberg2020-07-2743-2191/+3490
|\
| * [SERVER] Fix: NULL pointer access in saveLoadAllCacheMaps()Simon Rettberg2020-07-211-0/+2
| | | | | | | | Entries in _images array might ne NULL
| * [SERVER] Fix: No replication if autoFreeDiskSpaceDelay is disabledSimon Rettberg2020-06-301-4/+12
| |
| * [SHARED] Fix 16 byte information leakage in select image messageSimon Rettberg2020-06-301-2/+0Star
| |
| * get-version: Fallback to commit hash if git describe failsSimon Rettberg2020-06-161-1/+3
| |
| * [SERVER] Know when to stopSimon Rettberg2020-06-101-1/+1
| |
| * [SERVER] Likewise, flush entire payload on RTT measurementSimon Rettberg2020-06-101-1/+6
| |
| * [SERVER] Fix list walk when removing client from uplinkSimon Rettberg2020-06-101-1/+3
| |
| * [SERVER] fix uninitialized variableSimon Rettberg2020-06-101-1/+1
| |
| * [SERVER] Fix check for RTT block reply sizeSimon Rettberg2020-06-101-1/+1
| |
| * [SERVER] Check local and remote for updates on rid == 0Simon Rettberg2020-03-311-4/+8
| |
| * [BENCH] fix wrong operator precedenceSimon Rettberg2020-03-231-1/+1
| |
| * [SERVER] Make lock/thread debugging dedicated cmake optionSimon Rettberg2020-03-233-11/+19
| |
| * [SHARED] connect_ms might change above, don't checkSimon Rettberg2020-03-231-1/+1
| |
| * [SERVER] image_ensureDiskSpace should only deletes proxied imagesSimon Rettberg2020-03-201-18/+19
| |
| * [SERVER] Remember atime in .meta fileSimon Rettberg2020-03-203-62/+139
| |
| * [SERVER] Forbid hidden files when scanning image dirSimon Rettberg2020-03-201-1/+2
| |
| * [SERVER] Fix warnings, add assertionsSimon Rettberg2020-03-203-9/+15
| |
| * [SHARED] crc32: Don't skip table lookup if PCLMUL is unavailableSimon Rettberg2020-03-191-12/+13
| |
| * [SERVER] crc32: Fix compile with optimizationsSimon Rettberg2020-03-191-2/+2
| | | | | | | | | | Should have tested in "Release" mode I guess. Seems we're at about 24x performance this way, so hooray.
| * [SERVER] Use PCLMUL for crc32 on AMD64 if availableSimon Rettberg2020-03-192-55/+178
| | | | | | | | | | | | | | This is about 16x as fast as before with the lookup table for processing 4 bytes at a time and should work on any AMD64 CPU made in the last decade. We still need an AltiVec implementation for G5 though.
| * [SERVER] Shorter wait when closing socket after replySimon Rettberg2020-03-191-0/+2
| |
| * cmake: Fix version string dependency hell stuff once againSimon Rettberg2020-03-191-2/+3
| | | | | | | | | | BULLSHIT BULLSHIT
| * [SERVER] Add name param to threadpool_runSimon Rettberg2020-03-196-7/+16
| |
| * cmake: Fix version.c building on fresh checkoutsSimon Rettberg2020-03-181-4/+4
| |
| * [SERVER] Use first request in queue for RTT measurementsSimon Rettberg2020-03-171-6/+20
| | | | | | | | | | | | This makes sure the server we're potentially switching to can at least serve the next request in the queue, making sure some progress will be made.
| * [SERVER] Make sure bgrWindowSize doesn't overwhelm uplink queueSimon Rettberg2020-03-172-0/+6
| |
| * [SERVER] Include build type and version in RPCSimon Rettberg2020-03-176-13/+20
| | | | | | | | Added new query type q=version, which uses the STATS access rights.
| * [SERVER] Don't prefetch across hash blocks in BGS_HASHBLOCK modeSimon Rettberg2020-03-171-5/+8
| |
| * [SERVER] threadpool: Simplify get code, make debug code _DEBUG onlySimon Rettberg2020-03-141-4/+7
| |
| * [SERVER] Remove uplink_ prefix from static (private) functionsSimon Rettberg2020-03-141-40/+40
| |
| * [SERVER] Add comments, assert for uplink threadSimon Rettberg2020-03-141-6/+20
| |
| * [SERVER] Use image:rid in log messagesSimon Rettberg2020-03-141-5/+5
| |
| * [SERVER] Check server version before setting hopCount fieldSimon Rettberg2020-03-131-1/+1
| |
| * [SERVER] Check and increase hopCount when adding uplink requestSimon Rettberg2020-03-131-4/+7
| |
| * [SERVER] Rewrite uplink queue handlingSimon Rettberg2020-03-139-395/+518
| | | | | | | | | | | | - Now uses linked lists instead of huge array - Does prefetch data on client requests - Can have multiple replication requests in-flight
| * [SERVER] Fix data typeSimon Rettberg2020-03-091-2/+2
| |
| * [SERVER] Fix: Image would be assumed complete if no uplink existsSimon Rettberg2020-03-061-1/+1
| | | | | | | | Severe data corruption on client. Nice.
| * [SERVER] Add printf macro for image (name:rid as %s:%d)Simon Rettberg2020-03-064-30/+27Star
| |
| * [SERVER] Handle "warn unused result" casesSimon Rettberg2020-03-066-10/+26
| |
generated by cgit v1.2.3-55-g7522 (git 2.39.0) at 2024-06-27 20:16:02 +0200