From: Markus Schneider-Pargmann <msp@baylibre.com>
To: Chandrasekar Ramakrishnan <rcsekar@samsung.com>,
Marc Kleine-Budde <mkl@pengutronix.de>,
Wolfgang Grandegger <wg@grandegger.com>
Cc: linux-can@vger.kernel.org, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org,
Markus Schneider-Pargmann <msp@baylibre.com>
Subject: [PATCH 03/15] can: m_can: Cache tx putidx and transmits in flight
Date: Wed, 16 Nov 2022 21:52:56 +0100 [thread overview]
Message-ID: <20221116205308.2996556-4-msp@baylibre.com> (raw)
In-Reply-To: <20221116205308.2996556-1-msp@baylibre.com>
On peripheral chips every read/write can be costly. Avoid reading easily
trackable information and cache them internally. This saves multiple
reads.
Transmit FIFO put index is cached, this is increased for every time we
enqueue a transmit request.
The transmits in flight is cached as well. With each transmit request it
is increased when reading the finished transmit event it is decreased.
A submit limit is cached to avoid submitting too many transmits at once,
either because the TX FIFO or the TXE FIFO is limited. This is currently
done very conservatively as the minimum of the fifo sizes. This means we
can reach FIFO full events but won't drop anything.
Signed-off-by: Markus Schneider-Pargmann <msp@baylibre.com>
---
drivers/net/can/m_can/m_can.c | 21 +++++++++++++++------
drivers/net/can/m_can/m_can.h | 5 +++++
2 files changed, 20 insertions(+), 6 deletions(-)
diff --git a/drivers/net/can/m_can/m_can.c b/drivers/net/can/m_can/m_can.c
index 4adf03111782..f5bba848bd56 100644
--- a/drivers/net/can/m_can/m_can.c
+++ b/drivers/net/can/m_can/m_can.c
@@ -1041,6 +1041,7 @@ static int m_can_echo_tx_event(struct net_device *dev)
/* ack txe element */
m_can_write(cdev, M_CAN_TXEFA, FIELD_PREP(TXEFA_EFAI_MASK,
fgi));
+ --cdev->tx_fifo_in_flight;
/* update stats */
m_can_tx_update_stats(cdev, msg_mark, timestamp);
@@ -1376,6 +1377,14 @@ static void m_can_start(struct net_device *dev)
cdev->can.state = CAN_STATE_ERROR_ACTIVE;
m_can_enable_all_interrupts(cdev);
+
+ if (cdev->version > 30) {
+ cdev->tx_fifo_putidx = FIELD_GET(TXFQS_TFQPI_MASK,
+ m_can_read(cdev, M_CAN_TXFQS));
+ cdev->tx_fifo_in_flight = 0;
+ cdev->tx_fifo_submit_limit = min(cdev->mcfg[MRAM_TXE].num,
+ cdev->mcfg[MRAM_TXB].num);
+ }
}
static int m_can_set_mode(struct net_device *dev, enum can_mode mode)
@@ -1589,7 +1598,6 @@ static netdev_tx_t m_can_tx_handler(struct m_can_classdev *cdev)
struct sk_buff *skb = cdev->tx_skb;
struct id_and_dlc fifo_header;
u32 cccr, fdflags;
- u32 txfqs;
int err;
int putidx;
@@ -1646,10 +1654,8 @@ static netdev_tx_t m_can_tx_handler(struct m_can_classdev *cdev)
} else {
/* Transmit routine for version >= v3.1.x */
- txfqs = m_can_read(cdev, M_CAN_TXFQS);
-
/* Check if FIFO full */
- if (_m_can_tx_fifo_full(txfqs)) {
+ if (cdev->tx_fifo_in_flight >= cdev->tx_fifo_submit_limit) {
/* This shouldn't happen */
netif_stop_queue(dev);
netdev_warn(dev,
@@ -1665,7 +1671,7 @@ static netdev_tx_t m_can_tx_handler(struct m_can_classdev *cdev)
}
/* get put index for frame */
- putidx = FIELD_GET(TXFQS_TFQPI_MASK, txfqs);
+ putidx = cdev->tx_fifo_putidx;
/* Construct DLC Field, with CAN-FD configuration.
* Use the put index of the fifo as the message marker,
@@ -1699,9 +1705,12 @@ static netdev_tx_t m_can_tx_handler(struct m_can_classdev *cdev)
/* Enable TX FIFO element to start transfer */
m_can_write(cdev, M_CAN_TXBAR, (1 << putidx));
+ ++cdev->tx_fifo_in_flight;
+ cdev->tx_fifo_putidx = (++cdev->tx_fifo_putidx >= cdev->can.echo_skb_max ?
+ 0 : cdev->tx_fifo_putidx);
/* stop network queue if fifo full */
- if (m_can_tx_fifo_full(cdev) ||
+ if (cdev->tx_fifo_in_flight >= cdev->tx_fifo_submit_limit ||
m_can_next_echo_skb_occupied(dev, putidx))
netif_stop_queue(dev);
else if (cdev->is_peripheral && !cdev->tx_skb && netif_queue_stopped(dev))
diff --git a/drivers/net/can/m_can/m_can.h b/drivers/net/can/m_can/m_can.h
index 4c0267f9f297..7464ce56753a 100644
--- a/drivers/net/can/m_can/m_can.h
+++ b/drivers/net/can/m_can/m_can.h
@@ -92,6 +92,11 @@ struct m_can_classdev {
int pm_clock_support;
int is_peripheral;
+ // Store this internally to avoid fetch delays on peripheral chips
+ int tx_fifo_putidx;
+ int tx_fifo_in_flight;
+ int tx_fifo_submit_limit;
+
struct mram_cfg mcfg[MRAM_CFG_NUM];
};
--
2.38.1
next prev parent reply other threads:[~2022-11-16 20:54 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-16 20:52 [PATCH 00/15] can: m_can: Optimizations for tcan and peripheral chips Markus Schneider-Pargmann
2022-11-16 20:52 ` [PATCH 01/15] can: m_can: Eliminate double read of TXFQS in tx_handler Markus Schneider-Pargmann
2022-11-16 20:52 ` [PATCH 02/15] can: m_can: Wakeup net queue once tx was issued Markus Schneider-Pargmann
2022-11-30 17:21 ` Marc Kleine-Budde
2022-12-01 8:43 ` Markus Schneider-Pargmann
2022-12-01 9:16 ` Marc Kleine-Budde
2022-12-01 16:49 ` Markus Schneider-Pargmann
2022-12-02 9:16 ` Marc Kleine-Budde
2022-12-14 9:14 ` Markus Schneider-Pargmann
2022-12-14 9:18 ` Marc Kleine-Budde
2022-12-14 9:22 ` Marc Kleine-Budde
2022-12-14 10:15 ` Vincent MAILHOL
2022-12-14 10:35 ` Markus Schneider-Pargmann
2022-12-15 9:31 ` Markus Schneider-Pargmann
2022-12-16 4:40 ` Vincent MAILHOL
2022-12-14 10:18 ` Markus Schneider-Pargmann
2022-12-02 13:53 ` Marc Kleine-Budde
2022-11-16 20:52 ` Markus Schneider-Pargmann [this message]
2022-12-01 11:14 ` [PATCH 03/15] can: m_can: Cache tx putidx and transmits in flight Marc Kleine-Budde
2022-12-02 8:37 ` Markus Schneider-Pargmann
2022-12-02 14:46 ` Marc Kleine-Budde
2022-12-13 17:13 ` Markus Schneider-Pargmann
2022-12-13 19:17 ` Marc Kleine-Budde
2022-12-14 8:32 ` Markus Schneider-Pargmann
2022-11-16 20:52 ` [PATCH 04/15] can: m_can: Use transmit event FIFO watermark level interrupt Markus Schneider-Pargmann
2022-11-30 17:17 ` Marc Kleine-Budde
2022-12-01 8:25 ` Markus Schneider-Pargmann
2022-12-01 9:05 ` Marc Kleine-Budde
2022-12-01 10:12 ` Markus Schneider-Pargmann
2022-12-01 11:00 ` Marc Kleine-Budde
2022-12-01 16:59 ` Markus Schneider-Pargmann
2022-12-02 9:23 ` Marc Kleine-Budde
2022-12-02 9:43 ` Markus Schneider-Pargmann
2022-12-02 14:03 ` Marc Kleine-Budde
2022-12-13 17:19 ` Markus Schneider-Pargmann
2022-12-13 19:18 ` Marc Kleine-Budde
2022-11-16 20:52 ` [PATCH 05/15] can: m_can: Disable unused interrupts Markus Schneider-Pargmann
2022-11-16 20:52 ` [PATCH 06/15] can: m_can: Avoid reading irqstatus twice Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 07/15] can: m_can: Read register PSR only on error Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 08/15] can: m_can: Count TXE FIFO getidx in the driver Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 09/15] can: m_can: Count read getindex " Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 10/15] can: m_can: Batch acknowledge rx fifo Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 11/15] can: m_can: Batch acknowledge transmit events Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 12/15] can: tcan4x5x: Remove invalid write in clear_interrupts Markus Schneider-Pargmann
2022-12-02 14:17 ` Marc Kleine-Budde
2022-11-16 20:53 ` [PATCH 13/15] can: tcan4x5x: Fix use of register error status mask Markus Schneider-Pargmann
2022-12-02 14:19 ` Marc Kleine-Budde
2022-11-16 20:53 ` [PATCH 14/15] can: tcan4x5x: Fix register range of first block Markus Schneider-Pargmann
2022-12-02 14:28 ` Marc Kleine-Budde
2022-12-05 9:30 ` Markus Schneider-Pargmann
2022-12-05 9:44 ` Marc Kleine-Budde
2022-12-05 9:55 ` Markus Schneider-Pargmann
2022-11-16 20:53 ` [PATCH 15/15] can: tcan4x5x: Specify separate read/write ranges Markus Schneider-Pargmann
2022-12-02 14:03 ` [PATCH 00/15] can: m_can: Optimizations for tcan and peripheral chips Marc Kleine-Budde
2022-12-05 9:09 ` Markus Schneider-Pargmann
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=20221116205308.2996556-4-msp@baylibre.com \
--to=msp@baylibre.com \
--cc=linux-can@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mkl@pengutronix.de \
--cc=netdev@vger.kernel.org \
--cc=rcsekar@samsung.com \
--cc=wg@grandegger.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
all inboxes | Powered by JetHome®