lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <8B4A52CD-FC5A-4256-B7DE-A659B50654CE@amazon.com>
Date:   Mon, 2 Mar 2020 23:16:32 +0000
From:   "Machulsky, Zorik" <zorik@...zon.com>
To:     Josh Triplett <josh@...htriplett.org>,
        "Belgazal, Netanel" <netanel@...zon.com>,
        "Kiyanovski, Arthur" <akiyano@...zon.com>,
        "Tzalik, Guy" <gtzalik@...zon.com>,
        "Bshara, Saeed" <saeedb@...zon.com>
CC:     "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] ena: Speed up initialization 90x by reducing poll delays


On 2/28/20, 4:29 PM, "Josh Triplett" <josh@...htriplett.org> wrote:

    Before initializing completion queue interrupts, the ena driver uses
    polling to wait for responses on the admin command queue. The ena driver
    waits 5ms between polls, but the hardware has generally finished long
    before that. Reduce the poll time to 10us.
    
    On a c5.12xlarge, this improves ena initialization time from 173.6ms to
    1.920ms, an improvement of more than 90x. This improves server boot time
    and time to network bringup.
 
Thanks Josh,
We agree that polling rate should be increased, but prefer not to do it aggressively and blindly.
For example linear backoff approach might be a better choice. Please let us re-work a little this 
patch and bring it to review. Thanks!      
   
    Before:
    [    0.531722] calling  ena_init+0x0/0x63 @ 1
    [    0.531722] ena: Elastic Network Adapter (ENA) v2.1.0K
    [    0.531751] ena 0000:00:05.0: Elastic Network Adapter (ENA) v2.1.0K
    [    0.531946] PCI Interrupt Link [LNKD] enabled at IRQ 11
    [    0.547425] ena: ena device version: 0.10
    [    0.547427] ena: ena controller version: 0.0.1 implementation version 1
    [    0.709497] ena 0000:00:05.0: Elastic Network Adapter (ENA) found at mem febf4000, mac addr 06:c4:22:0e:dc:da, Placement policy: Low Latency
    [    0.709508] initcall ena_init+0x0/0x63 returned 0 after 173616 usecs
    
    After:
    [    0.526965] calling  ena_init+0x0/0x63 @ 1
    [    0.526966] ena: Elastic Network Adapter (ENA) v2.1.0K
    [    0.527056] ena 0000:00:05.0: Elastic Network Adapter (ENA) v2.1.0K
    [    0.527196] PCI Interrupt Link [LNKD] enabled at IRQ 11
    [    0.527211] ena: ena device version: 0.10
    [    0.527212] ena: ena controller version: 0.0.1 implementation version 1
    [    0.528925] ena 0000:00:05.0: Elastic Network Adapter (ENA) found at mem febf4000, mac addr 06:c4:22:0e:dc:da, Placement policy: Low Latency
    [    0.528934] initcall ena_init+0x0/0x63 returned 0 after 1920 usecs
    
    Signed-off-by: Josh Triplett <josh@...htriplett.org>
    ---
     drivers/net/ethernet/amazon/ena/ena_com.c | 22 ++++++++++++----------
     1 file changed, 12 insertions(+), 10 deletions(-)
    
    diff --git a/drivers/net/ethernet/amazon/ena/ena_com.c b/drivers/net/ethernet/amazon/ena/ena_com.c
    index 1fb58f9ad80b..203b2130d707 100644
    --- a/drivers/net/ethernet/amazon/ena/ena_com.c
    +++ b/drivers/net/ethernet/amazon/ena/ena_com.c
    @@ -62,7 +62,7 @@
     
     #define ENA_REGS_ADMIN_INTR_MASK 1
     
    -#define ENA_POLL_MS	5
    +#define ENA_POLL_US	10
     
     /*****************************************************************************/
     /*****************************************************************************/
    @@ -572,7 +572,7 @@ static int ena_com_wait_and_process_admin_cq_polling(struct ena_comp_ctx *comp_c
     			goto err;
     		}
     
    -		msleep(ENA_POLL_MS);
    +		usleep_range(ENA_POLL_US, 2 * ENA_POLL_US);
     	}
     
     	if (unlikely(comp_ctx->status == ENA_CMD_ABORTED)) {
    @@ -943,12 +943,13 @@ static void ena_com_io_queue_free(struct ena_com_dev *ena_dev,
     static int wait_for_reset_state(struct ena_com_dev *ena_dev, u32 timeout,
     				u16 exp_state)
     {
    -	u32 val, i;
    +	u32 val;
    +	unsigned long timeout_jiffies;
     
    -	/* Convert timeout from resolution of 100ms to ENA_POLL_MS */
    -	timeout = (timeout * 100) / ENA_POLL_MS;
    +	/* Convert timeout from resolution of 100ms */
    +	timeout_jiffies = jiffies + msecs_to_jiffies(timeout * 100);
     
    -	for (i = 0; i < timeout; i++) {
    +	while (1) {
     		val = ena_com_reg_bar_read32(ena_dev, ENA_REGS_DEV_STS_OFF);
     
     		if (unlikely(val == ENA_MMIO_READ_TIMEOUT)) {
    @@ -960,10 +961,11 @@ static int wait_for_reset_state(struct ena_com_dev *ena_dev, u32 timeout,
     			exp_state)
     			return 0;
     
    -		msleep(ENA_POLL_MS);
    -	}
    +		if (time_is_before_jiffies(timeout_jiffies))
    +			return -ETIME;
     
    -	return -ETIME;
    +		usleep_range(ENA_POLL_US, 2 * ENA_POLL_US);
    +	}
     }
     
     static bool ena_com_check_supported_feature_id(struct ena_com_dev *ena_dev,
    @@ -1458,7 +1460,7 @@ void ena_com_wait_for_abort_completion(struct ena_com_dev *ena_dev)
     	spin_lock_irqsave(&admin_queue->q_lock, flags);
     	while (atomic_read(&admin_queue->outstanding_cmds) != 0) {
     		spin_unlock_irqrestore(&admin_queue->q_lock, flags);
    -		msleep(ENA_POLL_MS);
    +		usleep_range(ENA_POLL_US, 2 * ENA_POLL_US);
     		spin_lock_irqsave(&admin_queue->q_lock, flags);
     	}
     	spin_unlock_irqrestore(&admin_queue->q_lock, flags);
    -- 
    2.25.1
    
    

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ