142 commits

Author	SHA1	Message	Date
Antonio SJ Musumeci	4fe0e90bc3	Fix dirent64::namelen calculation ... Worked fine on my machine where padding seemed to be all zeroed out but that doesn't appear to be the case elsewhere. Could try https://codeberg.org/jbruchon/libjodycode/src/branch/master/dir.c#L92-L110 but needs to be tested.	2025-09-30 11:01:31 -05:00
trapexit	46f85c0f6d	Misc fixes, mostly for FreeBSD (#1536)	2025-09-28 23:17:51 -05:00
trapexit	2f3e807e9d	Add getdents based readdir functions for Linux (#1533)	2025-09-23 23:51:21 -05:00
trapexit	ca59ae53a5	idmap mount support (#1523) ... * Add idmap mount support * Further replacement of std::filesystem::path to fs::path * Update docs regarding idmap support	2025-09-18 00:07:28 -05:00
trapexit	8c649d33ee	Start using std::filesystem::path for fusepath (#1522)	2025-09-17 22:51:16 -05:00
trapexit	071af2e516	More build process updates (#1516)	2025-08-29 07:51:05 -05:00
trapexit	b18aba86b6	Build improvements (#1501) ... * Remove python dependency * Remove ecfd * Improve makefiles * Make preload.so more 32bit compatible	2025-08-04 23:39:52 -05:00
trapexit	fa851eda3a	Ensure passthrough and keep_cache are mutually exclusive (#1493)	2025-07-06 18:31:47 -05:00
trapexit	f852474073	Update the secondary group cache (#1492) ... * Use concurrent_flat_map to build a global cache rather than per thread. * Create global maintanence thread manager for clearing unused cache entries.	2025-07-05 17:05:53 -05:00
trapexit	77ef88b04c	Add fsck.mergerfs tool (#1483)	2025-06-23 21:57:55 -05:00
trapexit	23824eaeb4	Misc cleanup	2025-06-06 19:10:19 -05:00
trapexit	e310d24615	Add support for FUSE passthrough	2025-06-05 23:47:42 -05:00
trapexit	92312c8507	Add debugging of mutexes (#1470) ... When DEBUG defined it will used timed locks which will log when held too long along with where.	2025-05-26 00:44:39 -05:00
trapexit	fa3266ca7f	Add podman release build tooling + misc build fixes (#1455)	2025-05-08 00:00:53 -05:00
trapexit	344014aa81	Rework thread pool for increased stability + config and doc updates (#1453) ... Also updates fmt to 11.2.0 and replaced syslog usage with a wrapper leveraging fmt.	2025-05-05 18:28:49 -05:00
Antonio SJ Musumeci	77bf16e9ae	Change the "devino" inode calculation ... Use hashing of branch path rather than use st_dev.	2025-04-15 18:54:51 -05:00
Antonio SJ Musumeci	80085c9844	Add statx support	2025-04-04 22:23:16 -05:00
Antonio SJ Musumeci	92ae63e842	Support Linux v6.13 FUSE max_page_limit	2025-03-30 00:07:08 -05:00
Antonio SJ Musumeci	b83bd5fd14	Improve FreeBSD compatibility ... First draft. Need to investigate why upstream patches include certain function that appear to exist in FreeBSD 14.2.	2025-01-15 22:10:02 -06:00
trapexit	0197984db3	Send invalidate node request outside lock	2025-01-09 23:22:07 -06:00
Antonio SJ Musumeci	9056898bbf	Move fuse.c and fuse_lowlevel.c to C++	2025-01-09 19:05:32 -06:00
Antonio SJ Musumeci	f0444a1ca9	Add support for 'direct-io-allow-mmap' if supported by kernel	2024-03-21 20:48:42 -05:00
Antonio SJ Musumeci	951eb9540d	Ensure lookups of FUSE_ROOT_ID always return generation=0	2024-02-25 17:14:15 -06:00
Antonio SJ Musumeci	0ca5d6aad0	Add export-support option as possible workaround for NFS EIO issues	2024-02-19 18:58:50 -06:00
Antonio SJ Musumeci	17fdbf7f42	Update fuse_kernel.h	2024-02-18 17:50:34 -06:00
Antonio SJ Musumeci	d6a2f06797	Only set allow_other when running as root	2023-10-29 01:44:18 -05:00
Antonio SJ Musumeci	6aa6452d3e	Add flushonclose feature	2023-10-28 19:47:23 -05:00
Antonio SJ Musumeci	fd33df04a3	Update wyhash to 4.2	2023-10-21 18:03:05 -05:00
Antonio SJ Musumeci	569537df9e	Remove use of pthread_getname_np to work with older musl versions	2023-10-14 01:15:29 -05:00
Antonio SJ Musumeci	620cab2948	Rework thread pool queue depth impl	2023-10-14 00:58:10 -05:00
Antonio SJ Musumeci	766b923116	Fix thread pool destruction where threads don't explicitly exit themselves	2023-09-09 16:48:47 -05:00
Antonio SJ Musumeci	6d2c99e487	Remove unneeded debug logging from thread pool	2023-09-04 18:39:43 -05:00
Antonio SJ Musumeci	0a94bd7cfc	Rework thread pool, add ability to add/remove threads at runtime	2023-09-03 23:36:03 -05:00
Antonio SJ Musumeci	76c8d48dbd	Change threadpool to use concurrentqueue	2023-09-02 00:45:47 -05:00
Antonio SJ Musumeci	0ed03a1535	Use relaxed memory order for atomic counters	2023-08-21 15:43:03 -05:00
Antonio SJ Musumeci	2377cb05b6	Add thread names for easier debugging	2023-08-14 17:51:41 -05:00
Antonio SJ Musumeci	0aafdefc18	Add ability to invalidate gid cache on demand ... SIGUSR2 or ioctl	2023-08-13 23:30:52 -05:00
Antonio SJ Musumeci	6dcf6111af	Add callbacks for newer functions ... * setupmapping * removemapping * syncfs * tmpfile	2023-08-13 17:38:13 -05:00
Antonio SJ Musumeci	eb141c4c4e	Fix logging pin-threads details	2023-07-30 23:13:14 -05:00
Antonio SJ Musumeci	51d97bb444	Move fuse thread args out of fuse session object	2023-07-30 22:52:43 -05:00
trapexit	36a4b7a23a	Merge pull request #1209 from trapexit/o_direct ... Align msg buffer memory to allow O_DIRECT to work	2023-07-09 19:39:52 -04:00
Antonio SJ Musumeci	22833bdfe2	Align msg buffer memory to allow O_DIRECT to work	2023-07-09 15:32:11 -05:00
Antonio SJ Musumeci	fa537961c5	Simplify syslog wrapper a bit	2023-07-06 00:13:42 -05:00
Antonio SJ Musumeci	5d6c9cc56a	Fix build on older distros due to std::atomic definitions ... std::atomic_uint64_t missing in some older distros	2023-07-02 23:46:46 -05:00
Antonio SJ Musumeci	5ab0fbcaee	Add manual GC triggering + configurable process queue depth ... Yes, these are unrelated changes but somehow ended up being prototyped together and I'm too lazy to separate them.	2023-06-26 22:44:23 -05:00
Antonio SJ Musumeci	6a14a10e6c	Fix read/write behavior and return value depending on direct_io ... Also add parallel direct write option for 6.2+ kernels.	2023-06-24 23:06:14 -05:00
Antonio SJ Musumeci	6bc3d77992	Fix crash when exiting with open deleted files	2023-04-05 00:32:30 -04:00
Antonio SJ Musumeci	373d331f39	Rework fuse read to use same buffers as messages	2023-02-26 23:31:44 -05:00
Antonio SJ Musumeci	84592a9f13	Remove splicing features ... After numerous tests it was found the splice features were at best the same performance as standard IO and at worse actually slower. To simplify the code all splice features are removed.	2023-02-26 01:42:58 -05:00
Antonio SJ Musumeci	34310170ac	Add ability to pin read and processing threads	2023-02-19 17:52:29 -05:00