From: Saeed Mahameed <saeed@kernel.org>
To: "David S. Miller" <davem@davemloft.net>,
Jakub Kicinski <kuba@kernel.org>
Cc: netdev@vger.kernel.org, Yevgeny Kliteynik <kliteyn@nvidia.com>,
Alex Vesker <valex@nvidia.com>,
Saeed Mahameed <saeedm@nvidia.com>
Subject: [net 05/19] net/mlx5: DR, Fix the threshold that defines when pool sync is initiated
Date: Wed, 23 Feb 2022 09:04:16 -0800 [thread overview]
Message-ID: <20220223170430.295595-6-saeed@kernel.org> (raw)
In-Reply-To: <20220223170430.295595-1-saeed@kernel.org>
From: Yevgeny Kliteynik <kliteyn@nvidia.com>
When deciding whether to start syncing and actually free all the "hot"
ICM chunks, we need to consider the type of the ICM chunks that we're
dealing with. For instance, the amount of available ICM for MODIFY_ACTION
is significantly lower than the usual STE ICM, so the threshold should
account for that - otherwise we can deplete MODIFY_ACTION memory just by
creating and deleting the same modify header action in a continuous loop.
This patch replaces the hard-coded threshold with a dynamic value.
Fixes: 1c58651412bb ("net/mlx5: DR, ICM memory pools sync optimization")
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
---
.../mellanox/mlx5/core/steering/dr_icm_pool.c | 11 +++++++----
1 file changed, 7 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_icm_pool.c b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_icm_pool.c
index f496b7e9401b..e289cfdbce07 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_icm_pool.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/steering/dr_icm_pool.c
@@ -4,7 +4,6 @@
#include "dr_types.h"
#define DR_ICM_MODIFY_HDR_ALIGN_BASE 64
-#define DR_ICM_SYNC_THRESHOLD_POOL (64 * 1024 * 1024)
struct mlx5dr_icm_pool {
enum mlx5dr_icm_type icm_type;
@@ -324,10 +323,14 @@ dr_icm_chunk_create(struct mlx5dr_icm_pool *pool,
static bool dr_icm_pool_is_sync_required(struct mlx5dr_icm_pool *pool)
{
- if (pool->hot_memory_size > DR_ICM_SYNC_THRESHOLD_POOL)
- return true;
+ int allow_hot_size;
- return false;
+ /* sync when hot memory reaches half of the pool size */
+ allow_hot_size =
+ mlx5dr_icm_pool_chunk_size_to_byte(pool->max_log_chunk_sz,
+ pool->icm_type) / 2;
+
+ return pool->hot_memory_size > allow_hot_size;
}
static int dr_icm_pool_sync_all_buddy_pools(struct mlx5dr_icm_pool *pool)
--
2.35.1
next prev parent reply other threads:[~2022-02-23 17:04 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-02-23 17:04 [pull request][net 00/19] mlx5 fixes 2022-02-23 Saeed Mahameed
2022-02-23 17:04 ` [net 01/19] net/mlx5: Update the list of the PCI supported devices Saeed Mahameed
2022-02-23 17:04 ` [net 02/19] net/mlx5: DR, Cache STE shadow memory Saeed Mahameed
2022-02-23 17:04 ` [net 03/19] net/mlx5: DR, Fix slab-out-of-bounds in mlx5_cmd_dr_create_fte Saeed Mahameed
2022-02-23 17:04 ` [net 04/19] net/mlx5: DR, Don't allow match on IP w/o matching on full ethertype/ip_version Saeed Mahameed
2022-02-23 23:26 ` Jakub Kicinski
2022-02-23 23:35 ` Yevgeny Kliteynik
2022-02-23 17:04 ` Saeed Mahameed [this message]
2022-02-23 17:04 ` [net 06/19] net/mlx5: Update log_max_qp value to be 17 at most Saeed Mahameed
2022-02-23 17:04 ` [net 07/19] net/mlx5: Fix wrong limitation of metadata match on ecpf Saeed Mahameed
2022-02-23 17:04 ` [net 08/19] net/mlx5: Fix tc max supported prio for nic mode Saeed Mahameed
2022-02-23 17:04 ` [net 09/19] net/mlx5: Fix possible deadlock on rule deletion Saeed Mahameed
2022-02-23 17:04 ` [net 10/19] net/mlx5e: Fix wrong return value on ioctl EEPROM query failure Saeed Mahameed
2022-02-23 17:04 ` [net 11/19] net/mlx5e: kTLS, Use CHECKSUM_UNNECESSARY for device-offloaded packets Saeed Mahameed
2022-02-23 17:04 ` [net 12/19] net/mlx5e: TC, Reject rules with drop and modify hdr action Saeed Mahameed
2022-02-23 17:04 ` [net 13/19] net/mlx5e: TC, Reject rules with forward and drop actions Saeed Mahameed
2022-02-23 17:04 ` [net 14/19] net/mlx5e: TC, Skip redundant ct clear actions Saeed Mahameed
2022-02-23 17:04 ` [net 15/19] net/mlx5e: Add feature check for set fec counters Saeed Mahameed
2022-02-23 17:04 ` [net 16/19] net/mlx5e: Fix MPLSoUDP encap to use MPLS action information Saeed Mahameed
2022-02-23 17:04 ` [net 17/19] net/mlx5e: MPLSoUDP decap, fix check for unsupported matches Saeed Mahameed
2022-02-23 17:04 ` [net 18/19] net/mlx5e: Add missing increment of count Saeed Mahameed
2022-02-23 17:04 ` [net 19/19] net/mlx5e: Fix VF min/max rate parameters interchange mistake Saeed Mahameed
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=20220223170430.295595-6-saeed@kernel.org \
--to=saeed@kernel.org \
--cc=davem@davemloft.net \
--cc=kliteyn@nvidia.com \
--cc=kuba@kernel.org \
--cc=netdev@vger.kernel.org \
--cc=saeedm@nvidia.com \
--cc=valex@nvidia.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).