From: Vivek Goyal <vgoyal@redhat.com>
To: qemu-devel@nongnu.org, virtio-fs-list <virtio-fs@redhat.com>
Cc: jose.carlos.venegas.munoz@intel.com
Subject: [Virtio-fs] [PATCH] virtiofsd: Use --thread-pool-size=0 to mean no thread pool
Date: Thu, 5 Nov 2020 14:44:16 -0500 [thread overview]
Message-ID: <20201105194416.GA1384085@redhat.com> (raw)
Right now we create a thread pool and main thread hands over the request
to thread in thread pool to process. Number of threads in thread pool
can be managed by option --thread-pool-size.
There is a chance that in case of some workloads, we might get better
performance if we don't handover the request to a different thread
and process in the context of thread receiving the request.
To implement that, redefine the meaning of --thread-pool-size=0 to
mean that don't use a thread pool. Instead process the request in
the context of thread receiving request from the queue.
I can't think how --thread-pool-size=0 is useful and hence using
that. If it is already useful somehow, I could look at defining
a new option say "--no-thread-pool".
I think this patch will be used more as a debug help to do comparison
when it is more effecient to do not hand over the requests to a
thread pool.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
---
tools/virtiofsd/fuse_virtio.c | 33 ++++++++++++++++++++++++---------
1 file changed, 24 insertions(+), 9 deletions(-)
diff --git a/tools/virtiofsd/fuse_virtio.c b/tools/virtiofsd/fuse_virtio.c
index ff86f6d1ce..60aa7cd3e5 100644
--- a/tools/virtiofsd/fuse_virtio.c
+++ b/tools/virtiofsd/fuse_virtio.c
@@ -695,13 +695,17 @@ static void *fv_queue_thread(void *opaque)
struct VuDev *dev = &qi->virtio_dev->dev;
struct VuVirtq *q = vu_get_queue(dev, qi->qidx);
struct fuse_session *se = qi->virtio_dev->se;
- GThreadPool *pool;
-
- pool = g_thread_pool_new(fv_queue_worker, qi, se->thread_pool_size, FALSE,
- NULL);
- if (!pool) {
- fuse_log(FUSE_LOG_ERR, "%s: g_thread_pool_new failed\n", __func__);
- return NULL;
+ GThreadPool *pool = NULL;
+ GList *req_list = NULL;
+
+ if (se->thread_pool_size) {
+ fuse_log(FUSE_LOG_DEBUG, "%s: Creating thread pool for Queue %d\n", __func__, qi->qidx);
+ pool = g_thread_pool_new(fv_queue_worker, qi, se->thread_pool_size,
+ FALSE, NULL);
+ if (!pool) {
+ fuse_log(FUSE_LOG_ERR, "%s: g_thread_pool_new failed\n", __func__);
+ return NULL;
+ }
}
fuse_log(FUSE_LOG_INFO, "%s: Start for queue %d kick_fd %d\n", __func__,
@@ -780,14 +784,25 @@ static void *fv_queue_thread(void *opaque)
req->bad_in_num = bad_in_num;
req->bad_out_num = bad_out_num;
- g_thread_pool_push(pool, req, NULL);
+ if (!se->thread_pool_size)
+ req_list = g_list_prepend(req_list, req);
+ else
+ g_thread_pool_push(pool, req, NULL);
}
pthread_mutex_unlock(&qi->vq_lock);
pthread_rwlock_unlock(&qi->virtio_dev->vu_dispatch_rwlock);
+
+ // Process all the requests.
+ if (!se->thread_pool_size && req_list != NULL) {
+ g_list_foreach(req_list, fv_queue_worker, qi);
+ g_list_free(req_list);
+ req_list = NULL;
+ }
}
- g_thread_pool_free(pool, FALSE, TRUE);
+ if (pool)
+ g_thread_pool_free(pool, FALSE, TRUE);
return NULL;
}
--
2.25.4
WARNING: multiple messages have this Message-ID (diff)
From: Vivek Goyal <vgoyal@redhat.com>
To: qemu-devel@nongnu.org, virtio-fs-list <virtio-fs@redhat.com>
Cc: jose.carlos.venegas.munoz@intel.com,
"Dr. David Alan Gilbert" <dgilbert@redhat.com>,
Stefan Hajnoczi <stefanha@redhat.com>
Subject: [PATCH] virtiofsd: Use --thread-pool-size=0 to mean no thread pool
Date: Thu, 5 Nov 2020 14:44:16 -0500 [thread overview]
Message-ID: <20201105194416.GA1384085@redhat.com> (raw)
Right now we create a thread pool and main thread hands over the request
to thread in thread pool to process. Number of threads in thread pool
can be managed by option --thread-pool-size.
There is a chance that in case of some workloads, we might get better
performance if we don't handover the request to a different thread
and process in the context of thread receiving the request.
To implement that, redefine the meaning of --thread-pool-size=0 to
mean that don't use a thread pool. Instead process the request in
the context of thread receiving request from the queue.
I can't think how --thread-pool-size=0 is useful and hence using
that. If it is already useful somehow, I could look at defining
a new option say "--no-thread-pool".
I think this patch will be used more as a debug help to do comparison
when it is more effecient to do not hand over the requests to a
thread pool.
Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
---
tools/virtiofsd/fuse_virtio.c | 33 ++++++++++++++++++++++++---------
1 file changed, 24 insertions(+), 9 deletions(-)
diff --git a/tools/virtiofsd/fuse_virtio.c b/tools/virtiofsd/fuse_virtio.c
index ff86f6d1ce..60aa7cd3e5 100644
--- a/tools/virtiofsd/fuse_virtio.c
+++ b/tools/virtiofsd/fuse_virtio.c
@@ -695,13 +695,17 @@ static void *fv_queue_thread(void *opaque)
struct VuDev *dev = &qi->virtio_dev->dev;
struct VuVirtq *q = vu_get_queue(dev, qi->qidx);
struct fuse_session *se = qi->virtio_dev->se;
- GThreadPool *pool;
-
- pool = g_thread_pool_new(fv_queue_worker, qi, se->thread_pool_size, FALSE,
- NULL);
- if (!pool) {
- fuse_log(FUSE_LOG_ERR, "%s: g_thread_pool_new failed\n", __func__);
- return NULL;
+ GThreadPool *pool = NULL;
+ GList *req_list = NULL;
+
+ if (se->thread_pool_size) {
+ fuse_log(FUSE_LOG_DEBUG, "%s: Creating thread pool for Queue %d\n", __func__, qi->qidx);
+ pool = g_thread_pool_new(fv_queue_worker, qi, se->thread_pool_size,
+ FALSE, NULL);
+ if (!pool) {
+ fuse_log(FUSE_LOG_ERR, "%s: g_thread_pool_new failed\n", __func__);
+ return NULL;
+ }
}
fuse_log(FUSE_LOG_INFO, "%s: Start for queue %d kick_fd %d\n", __func__,
@@ -780,14 +784,25 @@ static void *fv_queue_thread(void *opaque)
req->bad_in_num = bad_in_num;
req->bad_out_num = bad_out_num;
- g_thread_pool_push(pool, req, NULL);
+ if (!se->thread_pool_size)
+ req_list = g_list_prepend(req_list, req);
+ else
+ g_thread_pool_push(pool, req, NULL);
}
pthread_mutex_unlock(&qi->vq_lock);
pthread_rwlock_unlock(&qi->virtio_dev->vu_dispatch_rwlock);
+
+ // Process all the requests.
+ if (!se->thread_pool_size && req_list != NULL) {
+ g_list_foreach(req_list, fv_queue_worker, qi);
+ g_list_free(req_list);
+ req_list = NULL;
+ }
}
- g_thread_pool_free(pool, FALSE, TRUE);
+ if (pool)
+ g_thread_pool_free(pool, FALSE, TRUE);
return NULL;
}
--
2.25.4
next reply other threads:[~2020-11-05 19:44 UTC|newest]
Thread overview: 36+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-05 19:44 Vivek Goyal [this message]
2020-11-05 19:44 ` [PATCH] virtiofsd: Use --thread-pool-size=0 to mean no thread pool Vivek Goyal
2020-11-05 19:52 ` [Virtio-fs] " Vivek Goyal
2020-11-05 19:52 ` Vivek Goyal
2020-11-06 20:33 ` Venegas Munoz, Jose Carlos
2020-11-06 20:33 ` Venegas Munoz, Jose Carlos
2020-11-06 22:35 ` Vivek Goyal
2020-11-06 22:35 ` Vivek Goyal
2020-11-09 10:02 ` Dr. David Alan Gilbert
2020-11-09 10:02 ` Dr. David Alan Gilbert
2020-11-09 14:39 ` Vivek Goyal
2020-11-09 14:39 ` Vivek Goyal
2020-11-12 9:06 ` Miklos Szeredi
2020-11-12 9:57 ` Miklos Szeredi
2020-11-18 20:19 ` Vivek Goyal
2020-11-12 11:00 ` Christian Schoenebeck
2020-11-12 11:00 ` Christian Schoenebeck
2020-11-12 13:01 ` Miklos Szeredi
2020-11-17 16:00 ` Venegas Munoz, Jose Carlos
2020-11-17 16:39 ` Christian Schoenebeck
2020-11-17 16:39 ` Christian Schoenebeck
2020-11-17 18:55 ` Vivek Goyal
2020-11-17 18:55 ` Vivek Goyal
2020-11-23 15:06 ` Venegas Munoz, Jose Carlos
2020-11-23 15:06 ` Venegas Munoz, Jose Carlos
2020-11-17 22:24 ` Vivek Goyal
2020-11-17 22:24 ` Vivek Goyal
2020-11-23 15:07 ` Venegas Munoz, Jose Carlos
2020-11-23 15:07 ` Venegas Munoz, Jose Carlos
2020-11-18 19:51 ` Vivek Goyal
2020-11-23 9:46 ` Miklos Szeredi
2020-11-23 9:46 ` Miklos Szeredi
2020-11-09 10:11 ` Dr. David Alan Gilbert
2020-11-09 10:11 ` Dr. David Alan Gilbert
2020-11-09 14:40 ` [Virtio-fs] " Vivek Goyal
2020-11-09 14:40 ` Vivek Goyal
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201105194416.GA1384085@redhat.com \
--to=vgoyal@redhat.com \
--cc=jose.carlos.venegas.munoz@intel.com \
--cc=qemu-devel@nongnu.org \
--cc=virtio-fs@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.