Skip to content

K
Kernel

openeuler / Kernel
接近 2 年 前同步成功

通知 8
Star 0
Fork 0
K
Kernel
切换分支/标签
  1. 23 8月, 2019 1 次提交
  3. 15 6月, 2019 1 次提交
  5. 04 5月, 2019 1 次提交
  7. 17 4月, 2019 1 次提交
  9. 15 11月, 2018 1 次提交
  11. 08 8月, 2018 1 次提交
  13. 28 4月, 2018 1 次提交
  15. 24 3月, 2018 1 次提交
  17. 27 1月, 2018 2 次提交
  19. 24 1月, 2018 1 次提交
  21. 22 11月, 2017 1 次提交
  23. 03 10月, 2017 2 次提交
  25. 20 4月, 2017 1 次提交
    • J
      i40e: reduce wait time for adminq command completion · 9e3f23f4
      由 Jacob Keller 提交于 
      When sending an adminq command, we wait for the command to complete in
a loop. This loop waits for an entire millisecond, when in practice the
adminq command is processed often much faster.

Change the loop to use i40e_usec_delay instead, and wait for 50 usecs
each time instead. This appears to be about the minimum time required,
based on some manual observation and testing.

The primary benefit of this change is reducing latency of various
operations in the PF driver, especially when related to having a large
number of VFs enabled.

For example, on Linux, when instantiating 128 VFs, the time to finish
the operation dropped from about 9 seconds down to under 6 seconds.
Additionally, the time it takes to finish a PF reset with 128 VFs
dropped from 5.1 seconds down to 0.7 seconds.

As the examples above show, a significant portion of the delay is wasted
waiting for admiqn operations which have already finished.

This patch shouldn't cause impact to functionality, as we still check
and keep waiting until the command does get processed. The only expected
change is an increase in CPU utilization as we now check for completion
far more times. However, in practice the commands appear to generally be
complete within the first delay window anyways.

Change-ID: If8af8388e100da0a14eaf9e1af3afadf73a958cf
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
      9e3f23f4
  27. 01 11月, 2016 1 次提交
  29. 02 5月, 2016 1 次提交
  31. 07 4月, 2016 2 次提交
  33. 18 2月, 2016 1 次提交
  35. 04 12月, 2015 1 次提交
  37. 14 10月, 2015 1 次提交
  39. 09 10月, 2015 1 次提交
  41. 08 10月, 2015 1 次提交
  43. 07 10月, 2015 1 次提交
  45. 04 10月, 2015 1 次提交
  47. 29 9月, 2015 1 次提交
  49. 18 9月, 2015 3 次提交
  51. 05 3月, 2015 1 次提交
  53. 03 3月, 2015 1 次提交
  55. 10 12月, 2014 2 次提交
  57. 06 12月, 2014 2 次提交
  59. 21 11月, 2014 1 次提交
    • M
      i40e: don't overload fields · 1001dc37
      由 Mitch Williams 提交于 
      Overloading the msg_size field in the arq_event_info struct is just a
bad idea. It leads to repeated bugs when the structure is used in a
loop, since the input value (buffer size) is overwritten by the output
value (actual message length).

Fix this by splitting the field into two and renaming to indicate the
actual function of each field.

Since the arq_event struct has now changed, we need to change the drivers
to support this. Note that we no longer need to initialize the buffer size
each time we go through a loop as this value is no longer destroyed by
arq processing.

In the process, we also fix a bug in i40evf_verify_api_ver where the
buffer size was not correctly reinitialized each time through the loop.

Change-ID: Ic7f9633cdd6f871f93e698dfb095e29c696f5581
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Acked-by: NShannon Nelson <shannon.nelson@intel.com>
Acked-by: NAshish Shah <ashish.n.shah@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>
      1001dc37
  61. 11 11月, 2014 1 次提交
  63. 27 8月, 2014 1 次提交
  65. 24 7月, 2014 1 次提交