DPDK patches and discussions
 help / color / mirror / Atom feed
From: Gaetan Rivet <grive@u256.net>
To: dev@dpdk.org
Cc: Hemant Agrawal <hemant.agrawal@nxp.com>,
	Sachin Saxena <sachin.saxena@oss.nxp.com>
Subject: [dpdk-dev] [PATCH v2] bus/dpaa: optimize device name parsing
Date: Mon,  9 Nov 2020 14:37:55 +0100	[thread overview]
Message-ID: <20201109133755.999895-1-grive@u256.net> (raw)
In-Reply-To: <20201109113147.962882-1-grive@u256.net>

Device name parsing is done on all buses during device iterations at
either EAL or ethdev levels.

When a bus implements device name parsing slowly, all iterations are
impacted. Efficient implementation is important.

The DPAA bus device name parsing has two issues: it allocates dynamic
memory and uses snprintf without a real need for it. Both can be
avoided, which improves the parsing performance.

The function is also simpler and shorter.

Signed-off-by: Gaetan Rivet <grive@u256.net>
---
 drivers/bus/dpaa/dpaa_bus.c | 68 +++++++++++++++----------------------
 1 file changed, 28 insertions(+), 40 deletions(-)

diff --git a/drivers/bus/dpaa/dpaa_bus.c b/drivers/bus/dpaa/dpaa_bus.c
index ece6a4c424..a006872fae 100644
--- a/drivers/bus/dpaa/dpaa_bus.c
+++ b/drivers/bus/dpaa/dpaa_bus.c
@@ -385,12 +385,10 @@ dpaa_portal_finish(void *arg)
 }
 
 static int
-rte_dpaa_bus_parse(const char *name, void *out_name)
+rte_dpaa_bus_parse(const char *name, void *out)
 {
-	int i, j;
-	int max_fman = 2, max_macs = 16;
-	char *dup_name;
-	char *sep = NULL;
+	unsigned int i, j;
+	size_t delta;
 
 	/* There are two ways of passing device name, with and without
 	 * separator. "dpaa_bus:fm1-mac3" with separator, and "fm1-mac3"
@@ -399,46 +397,36 @@ rte_dpaa_bus_parse(const char *name, void *out_name)
 	 */
 	DPAA_BUS_DEBUG("Parse device name (%s)", name);
 
-	/* Check for dpaa_bus:fm1-mac3 style */
-	dup_name = strdup(name);
-	sep = strchr(dup_name, ':');
-	if (!sep)
-		/* If not, check for name=fm1-mac3 style */
-		sep = strchr(dup_name, '=');
+	delta = 0;
+	if (strncmp(name, "dpaa_bus:", 9) == 0) {
+		delta = 9;
+	} else if (strncmp(name, "name=", 5) == 0) {
+		delta = 5;
+	}
 
-	if (sep)
-		/* jump over the seprator */
-		sep = (char *) (sep + 1);
-	else
-		sep = dup_name;
-
-	for (i = 0; i < max_fman; i++) {
-		for (j = 0; j < max_macs; j++) {
-			char fm_name[16];
-			snprintf(fm_name, 16, "fm%d-mac%d", i, j);
-			if (strcmp(fm_name, sep) == 0) {
-				if (out_name)
-					strcpy(out_name, sep);
-				free(dup_name);
-				return 0;
-			}
-		}
+	if (sscanf(&name[delta], "fm%u-mac%u", &i, &j) != 2 ||
+	    i >= 2 || j >= 16) {
+		return -EINVAL;
 	}
 
-	for (i = 0; i < RTE_LIBRTE_DPAA_MAX_CRYPTODEV; i++) {
-		char sec_name[16];
-
-		snprintf(sec_name, 16, "dpaa_sec-%d", i+1);
-		if (strcmp(sec_name, sep) == 0) {
-			if (out_name)
-				strcpy(out_name, sep);
-			free(dup_name);
-			return 0;
-		}
+	if (out != NULL) {
+		char *out_name = out;
+		const size_t max_name_len = sizeof("fm.-mac..") - 1;
+
+		/* Do not check for truncation, either name ends with
+		 * '\0' or the device name is followed by parameters and there
+		 * will be a ',' instead. Not copying past this comma is not an
+		 * error.
+		 */
+		strlcpy(out_name, &name[delta], max_name_len + 1);
+
+		/* Second digit of mac%u could instead be ','. */
+		if ((strlen(out_name) == max_name_len) &&
+		    out_name[max_name_len] == ',')
+			out_name[max_name_len] = '\0';
 	}
 
-	free(dup_name);
-	return -EINVAL;
+	return 0;
 }
 
 #define DPAA_DEV_PATH1 "/sys/devices/platform/soc/soc:fsl,dpaa"
-- 
2.29.2


  parent reply	other threads:[~2020-11-09 13:38 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-11-09 11:31 [dpdk-dev] [PATCH] " Gaetan Rivet
2020-11-09 11:44 ` Gaëtan Rivet
2020-11-09 11:58 ` David Marchand
2020-11-09 13:36   ` Gaëtan Rivet
2020-11-09 13:37 ` Gaetan Rivet [this message]
2021-01-15 11:16   ` [dpdk-dev] [PATCH v2] " Thomas Monjalon

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=20201109133755.999895-1-grive@u256.net \
    --to=grive@u256.net \
    --cc=dev@dpdk.org \
    --cc=hemant.agrawal@nxp.com \
    --cc=sachin.saxena@oss.nxp.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).