aboutsummaryrefslogtreecommitdiffstats
path: root/src/util/disk_cache.c
Commit message (Collapse)AuthorAgeFilesLines
* util/u_queue: add UTIL_QUEUE_INIT_SET_FULL_THREAD_AFFINITYMarek Olšák2018-10-061-1/+2
| | | | | Initial version discussed with Rob Clark under a different patch name. This approach leaves his driver unaffected.
* util: rename timestamp param in disk_cache_create()Timothy Arceri2018-10-031-4/+4
| | | | | | | | Only some drivers use a timestamp here. Others use things such as build-id, or even a combination of build-ids from Mesa and LLVM. Reviewed-by: Marek Olšák <[email protected]>
* Shorten u_queue namesMarek Olšák2018-07-041-1/+1
| | | | | | | | There is a 15-character limit for thread names shared by the queue name and process name. Shorten the thread name to make space for the process name. Reviewed-by: Timothy Arceri <[email protected]>
* disk cache: move path creation back to constructorTapani Pälli2018-02-091-86/+76
| | | | | | | | | | | This patch moves disk cache path and index creation back to the constructor which matches previous behavior. We still allow create to succeed without path so that cache can be used with callback functionality. Fixes: c95d3ed091 "disk cache: create cache even if path creation fails" Signed-off-by: Tapani Pälli <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* disk cache: add callback functionalityTapani Pälli2018-02-071-0/+48
| | | | | | | | | | | v2: add disk_cache_has_key, disk_cache_put_key support using blob cache (Nicolai, Jordan) v3: rename set_cb as put_cb to match existing naming (Timothy) Signed-off-by: Tapani Pälli <[email protected]> Reviewed-by: Jordan Justen <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* disk cache: initialize cache path and index only when usedTapani Pälli2018-02-071-49/+80
| | | | | | | | | | | | This patch makes disk_cache initialize path and index lazily so that we can utilize disk_cache without a path using callback functionality introduced by next patch. v2: unmap mmap and destroy queue only if index_mmap exists Signed-off-by: Tapani Pälli <[email protected]> Reviewed-by: Jordan Justen <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* mesa: include compiler.h in disk_cache.cBrian Paul2018-01-171-0/+1
| | | | | | Instead of indirect inclusion to get CPU_TO_LE32() macro. Reviewed-by: Nicolai Hähnle <[email protected]>
* util/disk_cache: Remove unneeded free() on always null stringVadym Shovkoplias2017-12-071-1/+0
| | | | | | | | | At this point dc_job->cache_item_metadata.keys always equals NULL, so call to free() is useless Fixes: b86ecea3446 ("util/disk_cache: write cache item metadata to disk") Signed-off-by: Vadym Shovkoplias <[email protected]> Reviewed-by: Eric Engestrom <[email protected]>
* util: Fix disk_cache index calculation on big endianMatt Turner2017-11-251-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | The cache-test test program attempts to create a collision (using key_a and key_a_collide) by making the first two bytes identical. The idea is fine -- the shader cache wants to use the first four characters of a SHA1 hex digest as the index. The following program unsigned char array[4] = {1, 2, 3, 4}; int *ptr = (int *)array; for (int i = 0; i < 4; i++) { printf("%02x", array[i]); } printf("\n"); printf("%08x\n", *ptr); prints 01020304 04030201 on little endian, and 01020304 01020304 on big endian. On big endian platforms reading the character array back as an int (as is done in disk_cache.c) does not yield the same results as reading the byte array. To get the first four characters of the SHA1 hex digest when we mask with CACHE_INDEX_KEY_MASK, we need to byte swap the int on big endian platforms. Bugzilla: https://bugs.freedesktop.org/103668 Bugzilla: https://bugs.gentoo.org/637060 Bugzilla: https://bugs.gentoo.org/636326 Fixes: 87ab26b2ab35 ("glsl: Add initial functions to implement an on-disk cache") Reviewed-by: Emil Velikov <[email protected]>
* disk_cache: Fix issue reading GLSL metadataJordan Justen2017-10-311-1/+1
| | | | | | | | | | | | | | This would cause the read of the metadata content to fail, which would prevent the linking from being skipped. Seen on Rocket League with i965 shader cache. Fixes: b86ecea3446e "util/disk_cache: write cache item metadata to disk" Cc: Timothy Arceri <[email protected]> Signed-off-by: Jordan Justen <[email protected]> Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* disk_cache: remove unnecessary NULL-pointer guardsNicolai Hähnle2017-10-101-4/+2
| | | | Reviewed-by: Marek Olšák <[email protected]>
* disk_cache: fix a memory leakNicolai Hähnle2017-10-101-0/+1
| | | | Reviewed-by: Marek Olšák <[email protected]>
* util/disk_cache: turn MESA_GLSL_CACHE_DISABLE into a booleanEric Engestrom2017-09-121-1/+2
| | | | | | | | | | Instead of setting based on set/unset, allow users to use boolean values. In the docs and tests, use `DISABLE=true` instead of `DISABLE=1` as it's clearer IMO. Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* disk_cache: make the thread queue resizable and low priorityMarek Olšák2017-09-071-6/+8
| | | | | Acked-by: Timothy Arceri <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* disk_cache: assert if a cache entries keys don't match mesaTimothy Arceri2017-08-281-1/+3
| | | | | | | | | In ef42423e7be9 I enabled the check for release builds however we still want to assert in debug builds in case of collisions or just general bugs with the key building/compare code. Otherwise it will just fail silently effectively disabling the cache. Reviewed-by: Eduardo Lima Mitev <[email protected]>
* util/disk_cache: write cache item metadata to diskTimothy Arceri2017-08-251-1/+87
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* glsl: pass shader source keys to the disk cacheTimothy Arceri2017-08-251-3/+5
| | | | | | | We don't actually write them to disk here. That will happen in the following commit. Reviewed-by: Nicolai Hähnle <[email protected]>
* disk_cache: enable limited hash collision detection in release buildsTimothy Arceri2017-08-251-16/+10
| | | | | | | It really doesn't cost us much and will stop strange crashes should the stars align. Reviewed-by: Nicolai Hähnle <[email protected]>
* util/disk_cache: rename mesa cache dir and introduce cache versioningTimothy Arceri2017-08-251-12/+36
| | | | | | | | | | | | | | | | | | Steam is already analysing cache items, unfortunatly we did not introduce a versioning mechanism for identifying structural changes to cache entries earlier so the only way to do so is to rename the cache directory. Since we are renaming it we take the opportunity to give the directory a more meaningful name. Adding a version field to the header of cache entries will help us to avoid having to rename the directory in future. Please note this is versioning for the internal structure of the entries as defined in disk_cache.{c,h} as opposed to the structure of the data provided to the disk cache by the GLSL compiler and the various driver backends. Reviewed-by: Nicolai Hähnle <[email protected]>
* util/u_queue: add an option to set the minimum thread priorityMarek Olšák2017-06-071-1/+1
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* util/disk_cache: add new driver_flags param to cache keysTimothy Arceri2017-05-231-1/+7
| | | | | | | | | This will be used for things such as adding driver specific environment variables to the key. Allowing us to set environment vars that change the shader and not have the driver ignore them if it finds existing shaders in the cache. Reviewed-by: Eduardo Lima Mitev <[email protected]>
* util/disk_cache: remove percentage based max cache limitTimothy Arceri2017-04-281-6/+2
| | | | | | | | | | | | | | | | The more I think about it the more this seems like a bad idea. When we were deleting old cache dirs this wasn't so bad as it was unlikely we would ever hit the actual limit before things were cleaned up. Now that we only start cleaning up old cache items once the limit is reached the a percentage based max cache limit is more risky. For the inital release of shader cache I think its better to stick to a more conservative cache limit, at least until we have some way of cleaning up the cache more aggressively. Cc: "17.1" <[email protected]> Reviewed-by: Michel Dänzer <[email protected]>
* disk_cache: use block size rather than file sizeTimothy Arceri2017-04-271-5/+11
| | | | | | | | | | | | | | | | The majority of cache files are less than 1kb this resulted in us greatly miscalculating the amount of disk space used by the cache. Using the number of blocks allocated to the file is more conservative and less likely to cause issues. This change will result in cache sizes being miscalculated further until old items added with the previous calculation have all been removed. However I don't see anyway around that, the previous patch should help limit that problem. Cc: "17.1" <[email protected]> Reviewed-and-Tested-by: Michel Dänzer <[email protected]>
* disk_cache: reduce default cache size to 5% of filesystemTimothy Arceri2017-04-271-2/+2
| | | | | | | | | | | Modern disks are extremely large and are only going to get bigger. Usage has shown frequent Mesa upgrades can result in the cache growing very fast i.e. wasting a lot of disk space unnecessarily. 5% seems like a more reasonable default. Cc: "17.1" <[email protected]> Acked-by: Michel Dänzer <[email protected]>
* util/disk_cache: don't deadlock on premature EOFGrazvydas Ignotas2017-03-251-17/+26
| | | | | | | | | | | | | | If we get EOF earlier than expected, the current read loops will deadlock. This may easily happen if the disk cache gets corrupted. Fix it by using a helper function that handles EOF. Steps to reproduce (on a build with asserts disabled): $ glxgears $ find ~/.cache/mesa/ -type f -exec truncate -s 0 '{}' \; $ glxgears # deadlock Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: write cache entry keys to file headerTimothy Arceri2017-03-241-6/+44
| | | | | | | | | | | | This can be used to deal with key hash collisions from different versions (should we find that to actually happen) and to find which mesa version produced the cache entry. V2: use blob created at cache creation. v3: remove left over var from v1. Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: hash pointer size and gpu name into cache keysGrazvydas Ignotas2017-03-241-26/+15
| | | | | | | | | | | | | | | | This allows to get rid of the arch and gpu name directories. v2: (Timothy Arceri) don't use an opaque data type to store pointer size and gpu name. v3: (Timothy Arceri) use blob to store driver keys just make sure to store null terminator for strings, and make sure blob is defined by disk_cache and not it's users. v4: (Timothy Arceri) fix typo, and make ptr_size a uint8_t. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* util/disk_cache: hash timestamps into the cache keysGrazvydas Ignotas2017-03-241-51/+26
| | | | | | | | | | | | | | | | | | | | | | | | Instead of using a directory, hash the timestamps into the cache keys themselves. Since there is no more timestamp directory, there is no more need for deleting the cache of other mesa versions and we rely on eviction to clean up the old cache entries. This solves the problem of using several incarnations of disk_cache at the same time, where one deletes a directory belonging to the other, like when both OpenGL and gallium nine are used simultaneously (or several different mesa installations). v2: using additional blob instead of trying to clone sha1 state v3: (Timothy Arceri) don't use an opaque data type to store timestamp. V4: (Timothy Arceri) use blob to store driver keys just make sure to store null terminator for strings, and make sure blob is defined by disk_cache and not it's users. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=100091
* util/disk_cache: use rand_xorshift128plus() to get our random intTimothy Arceri2017-03-231-8/+10
| | | | | | | | | | | Otherwise for apps that don't seed the regular rand() we will always remove old cache entries from the same dirs. V2: assume bits returned by rand are independent uniformly distributed bits and grab our hex value without taking the modulus of the whole value, this also fixes a bug where 'f' was always missing. Reviewed-by: Nicolai Hähnle <[email protected]>
* util/disk_cache: add a write helperGrazvydas Ignotas2017-03-211-14/+23
| | | | | | | | | | | | | Simplifies the write code a bit and handles EINTR. V2: (Timothy Arceri) Drop EINTR handling. To do it properly we would need a retry limit but it's probably best to just avoid trying to write if we hit EINTR and try again next time we see the program. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: use a helper to compute cache keysGrazvydas Ignotas2017-03-211-0/+7
| | | | | | | | | | This will allow to hash additional data into the cache keys or even change the hashing algorithm easily, should we decide to do so. v2: don't try to compute key (and crash) if cache is disabled Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: check rename resultGrazvydas Ignotas2017-03-201-2/+6
| | | | | | | | | I haven't seen this causing problems in practice, but for correctness we should also check if rename succeeded to avoid breaking accounting and leaving a .tmp file behind. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: delete .tmp if target existsGrazvydas Ignotas2017-03-201-1/+3
| | | | | | | | | | | At the time of target file check, .tmp file is already created and file lock is held, so we should remove the .tmp, like in other error paths. With this, piglit no longer leaves large amount of empty .tmp files behind, which waste directory entries and may interfere with eviction. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: fix stored_keys indexGrazvydas Ignotas2017-03-201-2/+2
| | | | | | | | | | | | | | It seems there is a bug because: - 20 bytes are compared, but only 1 byte stored_keys step is used - entries can overlap each other by 19 bytes - index_mmap is ~1.3M in size, but only first 64K is used With this fix for Deus Ex: - startup time (from launch to Feral logo): ~38s -> ~16s - disk_cache_has_key() hit rate: ~50% -> ~96% Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: pass predicate functions file stats directly (v4)Alan Swanson2017-03-181-34/+21
| | | | | | | | | | | | | | | Since switching to LRU eviction the only user of these predicate functions now resolves directory entry stats itself so pass them directly saving calling fstat and strlen twice (and the expensive strlen is skipped entirely if access time is newer). v2: Update for empty cache dir detection changes v3: Fix passing string length to predicate with the +1 for NULL termination and also pass sb as pointer v4: Missed ampersand for passing sb as pointer Reviewed-by: Grazvydas Ignotas <[email protected]> Acked-by: Timothy Arceri <[email protected]>
* util/disk_cache: do eviction before creating .tmpGrazvydas Ignotas2017-03-161-10/+8
| | | | | | | | | | | | | | cache_put() first creates a .tmp file and then tries to do eviction. The recently added LRU eviction code selects non-empty directory with the oldest access time, but that may easily be the one with just the new .tmp file, especially on Linux where atime is updated lazily (with "relatime" mount option, which is the default). So when cache is small, if random doesn't hit another dir LRU keeps selecting the same dir with just the .tmp and not deleting anything. To fix this (and the tests), do eviction earlier. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: scale cache according to filesystem sizeAlan Swanson2017-03-151-3/+8
| | | | | | | | Select higher of current 1G default or 10% of filesystem where cache is located. Acked-by: Timothy Arceri <[email protected]> Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: actually enforce cache sizeAlan Swanson2017-03-151-1/+5
| | | | | | | | | | Currently only a one in one out eviction so if at max_size and cache files were to constantly increase in size then so would the cache. Restrict to limit of 8 evictions per new cache entry. V2: (Timothy Arceri) fix make check tests Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: use LRU eviction rather than random evictionAlan Swanson2017-03-151-43/+34
| | | | | | | | | | | Still using fast random selection of two-character subdirectory in which to check cache files rather than scanning entire cache. v2: Factor out double strlen call v3: C99 declaration of variables where used Reviewed-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: don't fallback to an empty cache dir on evictTimothy Arceri2017-03-151-6/+27
| | | | | | | | | | If we fail to randomly select a two letter cache dir, don't select an empty dir on fallback. In real world use we should never hit the fallback path but it can be hit by tests when the cache is set to a very small max value. Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: use a thread queue to write to shader cacheTimothy Arceri2017-03-151-13/+28
| | | | | | | | | | | | | | This should help reduce any overhead added by the shader cache when programs are not found in the cache. To avoid creating any special function just for the sake of the tests we add a one second delay whenever we call dick_cache_put() to give it time to finish. V2: poll for file when waiting for thread in test V3: fix poll delay to really be 100ms, and simplify the wait function Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: add helpers for creating/destroying disk cache put jobsTimothy Arceri2017-03-151-0/+40
| | | | | | | V2: Make a copy of the data so we don't have to worry about it being freed before we are done compressing/writing. Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: add thread queue to disk cacheTimothy Arceri2017-03-151-1/+15
| | | | | Reviewed-by: Marek Olšák <[email protected]> Reviewed-by: Grazvydas Ignotas <[email protected]>
* util/disk_cache: fix size subtraction on 32bitGrazvydas Ignotas2017-03-091-3/+3
| | | | | | | | | | Negating size_t on 32bit produces a 32bit result. This was effectively adding values close to UINT_MAX to the cache size (the files are usually small) instead of intended subtraction. Fixes 'make check' disk_cache failures on 32bit. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: fix compressed size calculationGrazvydas Ignotas2017-03-091-1/+1
| | | | | | | | | It incorrectly doubles the size on each iteration. Fixes: 85a9b1b5 "util/disk_cache: compress individual cache entries" Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: fix make checkTimothy Arceri2017-03-061-17/+0
| | | | | Fixes make check after 11f0efec2e615f5233d which caused disk cache to create an additional directory.
* Revert "glsl: Switch to disable-by-default for the GLSL shader cache"Timothy Arceri2017-03-061-7/+0
| | | | | | | | | This reverts commit 0f60c6616e93cba72bff4fbfedb72a753ef78e05. Piglit and all games tested so far seem to be working without issue. This change will allow wide user testing and we can decided before the next release if we need to turn it off again. Reviewed-by: Marek Olšák <[email protected]>
* util/disk_cache: support caches for multiple architecturesTimothy Arceri2017-03-061-0/+26
| | | | | | | | | | | | | | Previously we were deleting the entire cache if a user switched between 32 and 64 bit applications. V2: make the check more generic, it should now work with any platform we are likely to support. V3: Use suggestion from Emil to make even more generic/fix issue with __ILP32__ not being declared on gcc for regular 32-bit builds. Tested-by: Grazvydas Ignotas <[email protected]> Tested-by: Dieter Nützel <[email protected]>
* util/disk_cache: mark read-only arguments constGrazvydas Ignotas2017-03-061-17/+18
| | | | | | | No functional changes. Signed-off-by: Grazvydas Ignotas <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* util/disk_cache: compress individual cache entriesTimothy Arceri2017-03-031-23/+150
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This reduces the cache size for Deus Ex from ~160M to ~30M for radeonsi (these numbers differ from Grigori's results below probably due to different graphics quality settings). I'm also seeing the following improvements in minimum fps in the Shadow of Mordor benchmark on an i5-6400 [email protected], with a HDD: no-cache: ~10fps with-cache-no-compression: ~15fps with-cache-and-compression: ~20fps Note: The with cache results are from the second run after closing and opening the game to avoid the in-memory cache. Since we mainly care about decompression I went with Z_BEST_COMPRESSION as suggested on irc by Steinar H. Gunderson who has benchmarked decompression speeds. Grigori Goronzy provided the following stats for Deus Ex: Mankind Divided start-up times on a Athlon X4 860k with a SSD: No Cache 215 sec Cold Cache zlib BEST_COMPRESSION 285 sec Warm Cache zlib BEST_COMPRESSION 33 sec Cold Cache zlib BEST_SPEED 264 sec Warm Cache zlib BEST_SPEED 33 sec Cold Cache no compression 266 sec Warm Cache no compression 34 sec The total cache size for that game is 48 MiB with BEST_COMPRESSION, 56 MiB with BEST_SPEED and 170 MiB with no compression. These numbers suggest that it may be ok to go with Z_BEST_SPEED but we should gather some actual decompression times before doing so. Other options might be to do the compression in a separate thread, this might allow us to use a higher compression algorithim such as LZMA. Reviewed-by: Grigori Goronzy <[email protected]> Acked-by: Marek Olšák <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-12 04:50:52 +0000