summaryrefslogtreecommitdiff
path: root/src/rmd_cache.c
diff options
context:
space:
mode:
authorVito Caputo <vcaputo@pengaru.com>2020-11-09 13:01:01 -0800
committerVito Caputo <vcaputo@pengaru.com>2020-11-09 13:01:01 -0800
commit42904eef310313eeed6f8d91d5306488fd016d2e (patch)
treea240c60973296f52c6d32391c0ff731360efdf59 /src/rmd_cache.c
parent1a06acc0e75eacb6f284999e7530b729b3cb261c (diff)
cache: introduce "syncer" thread for cache writers
This adds a new flag: --periodic-datasync-ms with a default of 100ms. When a new CacheFile is created for writing, and the datasync period is non-zero, a thread is created for the sole purpose of calling fdatasync() on the underlying fd in a loop separated by sleeps of the specified duration. The purpose of this is to prevent a large backlog of dirty buffers from accumulating until the operating system's normal background dirty sync kicks in. Depending on how the underlying filesystem and storage stack is configured, these bulk writebacks can result in very long stalls on a subsequent write operation. This is easily observed on ext4+lvm+dmcrypt setups, even using a simple test like `dd if=/dev/urandom of=testfile bs=8M`. Despite having plenty of available buffers, once ext4's internal journal fills while a bulk writeback is underway, dd's progress will completely stall until the entire writeback is completed, usually it's in the jbd2 wchan of do_get_write_access(). When this occurs during a recording by recordMyDesktop, the result is dropping frames for the entire duration of the stall. One thing recordMyDesktop could do to insulate from this is perform its own buffering of sampled frames at the YUV stage, with the cache writer consuming from this pool of buffered frames. Then when the cache writer gets stalled on jbd2/dmcrypt holdups, the buffer pool just grows instead of frames being dropped. I may explore this option in the future, but for now simply syncing regularly has been sufficient in my usage, as it keeps the storage subsystem more continuously utilized and spreads out the writebacks so they don't completely back up the journal.
Diffstat (limited to 'src/rmd_cache.c')
-rw-r--r--src/rmd_cache.c57
1 files changed, 51 insertions, 6 deletions
diff --git a/src/rmd_cache.c b/src/rmd_cache.c
index db7682e..9e42be3 100644
--- a/src/rmd_cache.c
+++ b/src/rmd_cache.c
@@ -28,18 +28,48 @@
#include "rmd_cache.h"
#include "rmd_specsfile.h"
+#include "rmd_threads.h"
#include "rmd_types.h"
-#include <sys/types.h>
-#include <unistd.h>
-#include <stdlib.h>
+#include <fcntl.h>
#include <stdio.h>
+#include <stdlib.h>
#include <string.h>
#include <sys/stat.h>
+#include <sys/types.h>
+#include <unistd.h>
#define CACHE_FILE_SIZE_LIMIT (500 * 1024 * 1024)
+/* periodic fdatasync thread for cache writers when fdatasyncing is enabled */
+static void * rmdCacheFileSyncer(CacheFile *file)
+{
+ struct timespec delay;
+ int fd;
+
+ assert(file);
+ assert(file->periodic_datasync_ms);
+
+ rmdThreadsSetName("rmdCacheSyncer");
+
+ delay.tv_sec = file->periodic_datasync_ms / 1000;
+ delay.tv_nsec = (file->periodic_datasync_ms - delay.tv_sec * 1000) * 1000000;
+
+ if (file->gzfp)
+ fd = file->gzfd;
+ else
+ fd = fileno(file->fp);
+
+ for (;;) {
+ nanosleep(&delay, NULL);
+ fdatasync(fd);
+ }
+
+ return NULL;
+}
+
+
/* open file @ path storing the file handles in *file,
* this doesn't store the new path since it's primarily
* for the purposes of opening new chapters for the same
@@ -48,10 +78,12 @@
static int _rmdCacheFileOpen(CacheFile *file, const char *path)
{
const char *modestr = "rb";
+ int flags = O_RDONLY;
assert(file);
if (file->mode == RMD_CACHE_FILE_MODE_WRITE) {
+ flags = O_CREAT|O_WRONLY;
modestr = "wb";
if (file->compressed)
@@ -59,7 +91,12 @@ static int _rmdCacheFileOpen(CacheFile *file, const char *path)
}
if (file->compressed) {
- file->gzfp = gzopen(path, modestr);
+ /* zlib doesn't expose a fileno() equivalent for the syncer */
+ file->gzfd = open(path, flags, S_IRUSR|S_IWUSR);
+ if (file->gzfd < 0)
+ return -1;
+
+ file->gzfp = gzdopen(file->gzfd, modestr);
} else {
file->fp = fopen(path, modestr);
}
@@ -69,6 +106,9 @@ static int _rmdCacheFileOpen(CacheFile *file, const char *path)
file->chapter_n_bytes = 0;
+ if (file->mode == RMD_CACHE_FILE_MODE_WRITE && file->periodic_datasync_ms)
+ pthread_create(&file->syncer_thread, NULL, (void *(*)(void *))rmdCacheFileSyncer, file);
+
return 0;
}
@@ -78,6 +118,11 @@ static int _rmdCacheFileClose(CacheFile *file)
{
assert(file);
+ if (file->mode == RMD_CACHE_FILE_MODE_WRITE && file->periodic_datasync_ms) {
+ pthread_cancel(file->syncer_thread);
+ pthread_join(file->syncer_thread, NULL);
+ }
+
/* TODO: return meaningful -errno on errors? */
if (file->gzfp) {
if (gzclose(file->gzfp) != Z_OK)
@@ -114,8 +159,8 @@ CacheFile * rmdCacheFileOpen(ProgData *pdata, const char *path, CacheFileMode mo
}
f->mode = mode;
- if (!pdata->args.zerocompression)
- f->compressed = 1;
+ f->compressed = !pdata->args.zerocompression;
+ f->periodic_datasync_ms = pdata->args.periodic_datasync_ms;
if (_rmdCacheFileOpen(f, path) < 0)
return NULL;
© All Rights Reserved