DPDK patches and discussions
 help / color / mirror / Atom feed
From: Xiaolong Ye <xiaolong.ye@intel.com>
To: Olivier Matz <olivier.matz@6wind.com>,
	Thomas Monjalon <thomas@monjalon.net>,
	Konstantin Ananyev <konstantin.ananyev@intel.com>
Cc: dev@dpdk.org, Xiaolong Ye <xiaolong.ye@intel.com>, stable@dpdk.org
Subject: [dpdk-dev] [PATCH 3/5] mbuf: fix free_space setting for dynamic field
Date: Sat, 13 Jun 2020 23:49:19 +0800
Message-ID: <20200613154922.42379-4-xiaolong.ye@intel.com> (raw)
In-Reply-To: <20200613154922.42379-1-xiaolong.ye@intel.com>

The value free_space[i] is used to save the size of biggest aligned
element that can fit in the zone, current implementation has one flaw,
for example, if user registers dynfield1 (size = 4, align = 4, req = 124)
first, the free_space would be as below after registration:

  0070: 08 08 08 08 08 08 08 08
  0078: 08 08 08 08 00 00 00 00

Then if user continues to register dynfield2 (size = 4, align = 4),
free_space would become:

  0070: 00 00 00 00 04 04 04 04
  0078: 04 04 04 04 00 00 00 00

Further request dynfield3 (size = 8, align = 8) would fail to register
due to alignment requirement can't be satisfied, though there is enough
space remained in mbuf.

This patch fixes above issue by saving alignment only in aligned zone,
after the fix, above registrations order can be satisfied, free_space
would be like:

After dynfield1 registration:

  0070: 08 08 08 08 08 08 08 08
  0078: 04 04 04 04 00 00 00 00

After dynfield2 registration:

  0070: 08 08 08 08 08 08 08 08
  0078: 00 00 00 00 00 00 00 00

After dynfield3 registration:

  0070: 00 00 00 00 00 00 00 00
  0078: 00 00 00 00 00 00 00 00

This patch also reduces iterations in process_score() by jumping align
steps in each loop.

Fixes: 4958ca3a443a ("mbuf: support dynamic fields and flags")
Cc: stable@dpdk.org

Signed-off-by: Xiaolong Ye <xiaolong.ye@intel.com>
---
 lib/librte_mbuf/rte_mbuf_dyn.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/lib/librte_mbuf/rte_mbuf_dyn.c b/lib/librte_mbuf/rte_mbuf_dyn.c
index de7d2eb9a5..fd51e1b68e 100644
--- a/lib/librte_mbuf/rte_mbuf_dyn.c
+++ b/lib/librte_mbuf/rte_mbuf_dyn.c
@@ -67,13 +67,16 @@ process_score(void)
 			shm->free_space[i] = 1;
 	}
 
-	for (off = 0; off < sizeof(struct rte_mbuf); off++) {
+	off = 0;
+	while (off < sizeof(struct rte_mbuf)) {
 		/* get the size of the free zone */
 		for (size = 0; (off + size) < sizeof(struct rte_mbuf) &&
 			     shm->free_space[off + size]; size++)
 			;
-		if (size == 0)
+		if (size == 0) {
+			off++;
 			continue;
+		}
 
 		/* get the alignment of biggest object that can fit in
 		 * the zone at this offset.
@@ -84,8 +87,10 @@ process_score(void)
 			;
 
 		/* save it in free_space[] */
-		for (i = off; i < off + size; i++)
+		for (i = off; i < off + align; i++)
 			shm->free_space[i] = RTE_MAX(align, shm->free_space[i]);
+
+		off += align;
 	}
 }
 
-- 
2.17.1


  parent reply	other threads:[~2020-06-13 15:58 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-13 15:49 [dpdk-dev] [PATCH 0/5] small fixes for mbuf's dynamic field/flag feature Xiaolong Ye
2020-06-13 15:49 ` [dpdk-dev] [PATCH 1/5] mbuf: fix out-of-bounds access at dyn field register Xiaolong Ye
2020-06-13 15:49 ` [dpdk-dev] [PATCH 2/5] mbuf: fix missing errno for dyn field/flag registration Xiaolong Ye
2020-06-25 15:31   ` Olivier Matz
2020-06-13 15:49 ` Xiaolong Ye [this message]
2020-06-25 15:53   ` [dpdk-dev] [PATCH 3/5] mbuf: fix free_space setting for dynamic field Olivier Matz
2020-06-13 15:49 ` [dpdk-dev] [PATCH 4/5] mbuf: fix a dynamic field dump log Xiaolong Ye
2020-06-25 15:54   ` Olivier Matz
2020-06-13 15:49 ` [dpdk-dev] [PATCH 5/5] mbuf: support to dump free_flags for dynamic flag Xiaolong Ye
2020-06-25 15:55   ` Olivier Matz
2020-06-25 20:40 ` [dpdk-dev] [PATCH 0/5] small fixes for mbuf's dynamic field/flag feature 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=20200613154922.42379-4-xiaolong.ye@intel.com \
    --to=xiaolong.ye@intel.com \
    --cc=dev@dpdk.org \
    --cc=konstantin.ananyev@intel.com \
    --cc=olivier.matz@6wind.com \
    --cc=stable@dpdk.org \
    --cc=thomas@monjalon.net \
    /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

DPDK patches and discussions

This inbox may be cloned and mirrored by anyone:

	git clone --mirror https://inbox.dpdk.org/dev/0 dev/git/0.git

	# If you have public-inbox 1.1+ installed, you may
	# initialize and index your mirror using the following commands:
	public-inbox-init -V2 dev dev/ https://inbox.dpdk.org/dev \
		dev@dpdk.org
	public-inbox-index dev

Example config snippet for mirrors.
Newsgroup available over NNTP:
	nntp://inbox.dpdk.org/inbox.dpdk.dev


AGPL code for this site: git clone https://public-inbox.org/public-inbox.git