[v4,00/21] Add perf support to the rockchip-dfi driver

Message

Sascha Hauer May 5, 2023, 11:38 a.m. UTC

This is v4 of the patches adding DDR perf support to the rockchip-dfi
driver.

There was no reaction to v3 of this series. Heiko, as the devfreq maintainers
are unreactive, would it be an option to take this through your tree?

Compared to the last version I added RK3588 support this time which required
some minor refactoring of the driver code. Only the driver changes are
included, not yet the device tree changes.

To make a bit of advertisement for this series: It allows us to measure the
DDR utilization which can be quite nice to identify performance bottlenecks.

Usage can be something like:

perf stat -a -e rockchip_ddr/cycles/,\
                rockchip_ddr/read-bytes/,\
                rockchip_ddr/write-bytes/,\
                rockchip_ddr/bytes/ sleep 1

 Performance counter stats for 'system wide':

        1582524826      rockchip_ddr/cycles/
           1802.25 MB   rockchip_ddr/read-bytes/
           1793.72 MB   rockchip_ddr/write-bytes/
           3595.90 MB   rockchip_ddr/bytes/

       1.014369709 seconds time elapsed

The above was called with running a test program in the background which copies
memory around. The measurements actually match quite well to the amount of
memory really copied as output by the test program.

Sascha

Changes since v3:
- Add RK3588 support

Changes since v2:
- Fix broken reference to binding
- Add Reviewed-by from Rob

Changes since v1:
- Fix example to actually match the binding and fix the warnings resulted thereof
- Make addition of rockchip,rk3568-dfi an extra patch

Sascha Hauer (21):
  PM / devfreq: rockchip-dfi: Embed desc into private data struct
  PM / devfreq: rockchip-dfi: use consistent name for private data
    struct
  PM / devfreq: rockchip-dfi: Make pmu regmap mandatory
  PM / devfreq: rockchip-dfi: Add SoC specific init function
  PM / devfreq: rockchip-dfi: dfi store raw values in counter struct
  PM / devfreq: rockchip-dfi: Use free running counter
  PM / devfreq: rockchip-dfi: introduce channel mask
  PM / devfreq: rk3399_dmc,dfi: generalize DDRTYPE defines
  PM / devfreq: rockchip-dfi: Clean up DDR type register defines
  PM / devfreq: rockchip-dfi: Add RK3568 support
  PM / devfreq: rockchip-dfi: Handle LPDDR2 correctly
  PM / devfreq: rockchip-dfi: Handle LPDDR4X
  PM / devfreq: rockchip-dfi: Pass private data struct to internal
    functions
  PM / devfreq: rockchip-dfi: Prepare for multiple users
  PM / devfreq: rockchip-dfi: Add perf support
  PM / devfreq: rockchip-dfi: make register stride SoC specific
  PM / devfreq: rockchip-dfi: account for multiple DDRMON_CTRL registers
  PM / devfreq: rockchip-dfi: add support for RK3588
  arm64: dts: rockchip: rk3399: Enable DFI
  arm64: dts: rockchip: rk356x: Add DFI
  dt-bindings: devfreq: event: convert Rockchip DFI binding to yaml

 .../bindings/devfreq/event/rockchip-dfi.txt   |  18 -
 .../bindings/devfreq/event/rockchip-dfi.yaml  |  39 +
 arch/arm64/boot/dts/rockchip/rk3399.dtsi      |   1 -
 arch/arm64/boot/dts/rockchip/rk356x.dtsi      |   6 +
 drivers/devfreq/event/rockchip-dfi.c          | 723 +++++++++++++++---
 drivers/devfreq/rk3399_dmc.c                  |  10 +-
 include/soc/rockchip/rk3399_grf.h             |   9 +-
 include/soc/rockchip/rk3568_grf.h             |  13 +
 include/soc/rockchip/rk3588_grf.h             |  18 +
 include/soc/rockchip/rockchip_grf.h           |  16 +
 10 files changed, 718 insertions(+), 135 deletions(-)
 delete mode 100644 Documentation/devicetree/bindings/devfreq/event/rockchip-dfi.txt
 create mode 100644 Documentation/devicetree/bindings/devfreq/event/rockchip-dfi.yaml
 create mode 100644 include/soc/rockchip/rk3568_grf.h
 create mode 100644 include/soc/rockchip/rk3588_grf.h
 create mode 100644 include/soc/rockchip/rockchip_grf.h

Comments

Vincent Legoll May 5, 2023, 4:38 p.m. UTC | #1

Hello,

On Fri, May 5, 2023 at 11:41 AM Sascha Hauer <[e-mail]> wrote:
> This is v4 of the patches adding DDR perf support to the rockchip-dfi
> driver.

I tested this series on a Pine64 QuartzPro64 dev board.

I applied the series on top of my branch, which is based on Collabora's
rockchip-3588 plus some QP64 DTS patches, and your WIP patch to
enable the series in the rk3588s.dtsi.

Looks like this is working properly:

# perf list | grep rockchip_ddr
  rockchip_ddr/bytes/                                [Kernel PMU event]
  rockchip_ddr/cycles/                               [Kernel PMU event]
  rockchip_ddr/read-bytes/                           [Kernel PMU event]
  rockchip_ddr/write-bytes/                          [Kernel PMU event]

# and while doing nothing...
# perf stat -a -e
rockchip_ddr/cycles/,rockchip_ddr/read-bytes/,rockchip_ddr/write-bytes/,rockchip_ddr/bytes/
sleep 1

 Performance counter stats for 'system wide':

        1058497566      rockchip_ddr/cycles/
              9.01 MB   rockchip_ddr/read-bytes/
              0.66 MB   rockchip_ddr/write-bytes/
              9.65 MB   rockchip_ddr/bytes/

       1.002460137 seconds time elapsed

# then with a background memory (ab-)user...
# memtester 4G > /dev/null 2>&1 &
# perf stat -a -e
rockchip_ddr/cycles/,rockchip_ddr/read-bytes/,rockchip_ddr/write-bytes/,rockchip_ddr/bytes/
sleep 10

 Performance counter stats for 'system wide':

       10561937833      rockchip_ddr/cycles/
          59192.90 MB   rockchip_ddr/read-bytes/
          29104.02 MB   rockchip_ddr/write-bytes/
          88296.92 MB   rockchip_ddr/bytes/

      10.001867723 seconds time elapsed

You can add my T-B, for the whole series:

Tested-by: Vincent Legoll <[e-mail]>

Thanks for your work
Regards

--
Vincent Legoll