From: Slavomir Kaslev <kaslevs@vmware.com>
To: linux-trace-devel@vger.kernel.org
Cc: rostedt@goodmis.org, ykaradzhov@vmware.com, tstoyanov@vmware.com
Subject: [PATCH v4 8/8] trace-cmd: Use splice(2) for vsock sockets if available
Date: Wed, 16 Jan 2019 15:43:07 +0200 [thread overview]
Message-ID: <20190116134307.4185-9-kaslevs@vmware.com> (raw)
In-Reply-To: <20190116134307.4185-1-kaslevs@vmware.com>
Detect if splice(2) reading is supported for vsock sockets (Linux 4.20 and
later) and use it, or fallback to read/write-ing otherwise.
Signed-off-by: Slavomir Kaslev <kaslevs@vmware.com>
---
tracecmd/trace-record.c | 59 ++++++++++++++++++++++++++++++++++++++---
1 file changed, 56 insertions(+), 3 deletions(-)
diff --git a/tracecmd/trace-record.c b/tracecmd/trace-record.c
index 1622b5e..cd80462 100644
--- a/tracecmd/trace-record.c
+++ b/tracecmd/trace-record.c
@@ -2658,12 +2658,61 @@ static int open_vsock(unsigned int cid, unsigned int port)
return sd;
}
+
+static int try_splice_read_vsock(void)
+{
+ int ret, sd, brass[2];
+
+ sd = socket(AF_VSOCK, SOCK_STREAM, 0);
+ if (sd < 0)
+ return sd;
+
+ ret = pipe(brass);
+ if (ret < 0)
+ goto out_close_sd;
+
+ /*
+ * On kernels that don't support splice reading from vsock sockets
+ * this will fail with EINVAL, or ENOTCONN otherwise.
+ * Technically, it should never succeed but if it does, claim splice
+ * reading is supported.
+ */
+ ret = splice(sd, NULL, brass[1], NULL, 10, 0);
+ if (ret < 0)
+ ret = errno != EINVAL;
+ else
+ ret = 1;
+
+ close(brass[0]);
+ close(brass[1]);
+out_close_sd:
+ close(sd);
+ return ret;
+}
+
+static bool can_splice_read_vsock(void)
+{
+ static bool initialized, res;
+
+ if (initialized)
+ return res;
+
+ res = try_splice_read_vsock() > 0;
+ initialized = true;
+ return res;
+}
+
#else
static inline int open_vsock(unsigned int cid, unsigned int port)
{
die("vsock is not supported");
return -1;
}
+
+static bool can_splice_read_vsock(void)
+{
+ return false;
+}
#endif
static int do_accept(int sd)
@@ -2868,13 +2917,16 @@ create_recorder_instance(struct buffer_instance *instance, const char *file, int
if (instance->flags & BUFFER_FL_GUEST) {
int sd;
+ unsigned int flags;
sd = open_vsock(instance->cid, instance->client_ports[cpu]);
if (sd < 0)
die("Failed to connect to agent");
- return tracecmd_create_recorder_virt(
- file, cpu, recorder_flags | TRACECMD_RECORD_NOSPLICE, sd);
+ flags = recorder_flags;
+ if (!can_splice_read_vsock())
+ flags |= TRACECMD_RECORD_NOSPLICE;
+ return tracecmd_create_recorder_virt(file, cpu, flags, sd);
}
if (brass)
@@ -2928,7 +2980,8 @@ static int create_recorder(struct buffer_instance *instance, int cpu,
if (instance->flags & BUFFER_FL_AGENT) {
fd = do_accept(instance->fds[cpu]);
- flags |= TRACECMD_RECORD_NOSPLICE;
+ if (!can_splice_read_vsock())
+ flags |= TRACECMD_RECORD_NOSPLICE;
} else {
fd = connect_port(host, instance->client_ports[cpu]);
}
--
2.19.1
prev parent reply other threads:[~2019-01-16 13:43 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-16 13:42 [PATCH v4 0/8] Add VM kernel tracing over vsock sockets Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 1/8] trace-cmd: Minor refactoring Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 2/8] trace-cmd: Detect if vsock sockets are available Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 3/8] trace-cmd: Add tracecmd_create_recorder_virt function Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 4/8] trace-cmd: Use unsigned int for trace-cmd client ports Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 5/8] trace-cmd: Add TRACE_REQ and TRACE_RESP messages Slavomir Kaslev
2019-01-16 14:19 ` Steven Rostedt
2019-01-16 13:43 ` [PATCH v4 6/8] trace-cmd: Add buffer instance flags for tracing in guest and agent context Slavomir Kaslev
2019-01-16 13:43 ` [PATCH v4 7/8] trace-cmd: Add VM kernel tracing over vsock sockets transport Slavomir Kaslev
2019-01-16 13:43 ` Slavomir Kaslev [this message]
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=20190116134307.4185-9-kaslevs@vmware.com \
--to=kaslevs@vmware.com \
--cc=linux-trace-devel@vger.kernel.org \
--cc=rostedt@goodmis.org \
--cc=tstoyanov@vmware.com \
--cc=ykaradzhov@vmware.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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).