Re: [v4 00/11] Improve RISC-V Perf support using SBI PMU and sscofpmf extension

From: Palmer Dabbelt
Date: Mon Dec 13 2021 - 20:51:39 EST

Next message: Masahiro Yamada: "Re: [PATCH 08/10] kbuild: do not include include/config/auto.conf from shell scripts"
Previous message: Atish Patra: "Re: [PATCH 1/2] riscv: prevent null-pointer dereference with sbi_remote_fence_i"
Next in thread: Atish Patra: "Re: [v4 00/11] Improve RISC-V Perf support using SBI PMU and sscofpmf extension"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Mon, 25 Oct 2021 12:53:39 PDT (-0700), Atish Patra wrote:

This series adds improved perf support for RISC-V based system using
SBI PMU extension[1] and Sscofpmf extension[2]. The SBI PMU extension allows
the kernel to program the counters for different events and start/stop counters
while the sscofpmf extension allows the counter overflow interrupt and privilege
mode filtering. An hardware platform can leverage SBI PMU extension without
the sscofpmf extension if it supports mcountinhibit and mcounteren. However,
the reverse is not true. With both of these extension enabled, a platform can
take advantage of all both event counting and sampling using perf tool.

This series introduces a platform perf driver instead of a existing arch
specific implementation. The new perf implementation has adopted a modular
approach where most of the generic event handling is done in the core library
while individual PMUs need to only implement necessary features specific to
the PMU. This is easily extensible and any future RISC-V PMU implementation
can leverage this. Currently, SBI PMU driver & legacy PMU driver are implemented
as a part of this series.

The legacy driver tries to reimplement the existing minimal perf under a new
config to maintain backward compatibility. This implementation only allows
monitoring of always running cycle/instruction counters. Moreover, they can
not be started or stopped. In general, this is very limited and not very useful.
That's why, I am not very keen to carry the support into the new driver.
However, I don't want to break perf for any existing hardware platforms.
If everybody agrees that we don't need legacy perf implementation for older
implementation, I will be happy to drop PATCH 4.

IMO we should keep it for a bit, so we have a transition period. These extensions are pretty new so we won't be able to count on everyone having them yet, this way we'll avoid breaking users.

This generally looks good, but I don't see any Acks from the perf maintainers. I'm happy to take this through the RISC-V tree, but I'd generally like to have at least an ack as perf isn't really my subsystem. MAINTAINERS seems to indicate that's Will and Mark, they're not To'd so maybe they just missed this?

I fixed a few trivial checkpatch warnings, updated Atish's email address, and put this on palmer/riscv-pmu. Happy to hear any comments, if nobody says anything then I'll just put that on riscv/for-next whenever I get back to my own email.

This series has been tested in Qemu (RV64 & RV32) and HiFive Unmatched.
Qemu[5] & OpenSBI [3] patches are required to test it on Qemu and a dt patch
required in U-Boot[6] for HiFive Unmatched. Qemu changes are not
backward compatible. That means, you can not use perf anymore on older Qemu
versions with latest OpenSBI and/or Kernel. However, newer kernel will
just use legacy pmu driver if old OpenSBI is detected.

The U-Boot patch is just an example that encodes few of the events defined
in fu740 documentation [7] in the DT. We can update the DT to include all the
events defined if required.

Here is an output of perf stat/report while running perf benchmark with OpenSBI,
Linux kernel and U-Boot patches applied.

HiFive Unmatched:
=================
perf stat -e cycles -e instructions -e L1-icache-load-misses -e branches -e branch-misses \
-e r0000000000000200 -e r0000000000000400 \
-e r0000000000000800 perf bench sched messaging -g 25 -l 15

# Running 'sched/messaging' benchmark:
# 20 sender and receiver processes per group
# 25 groups == 1000 processes run

Total time: 0.826 [sec]

Performance counter stats for 'perf bench sched messaging -g 25 -l 15':

3426710073 cycles (65.92%)
1348772808 instructions #0.39 insn per cycle (75.44%)
0 L1-icache-load-misses (72.28%)
201133996 branches (67.88%)
44663584 branch-misses #22.21% of all branches (35.01%)
248194747 r0000000000000200 (41.94%) --> Integer load instruction retired
156879950 r0000000000000400 (43.58%) --> Integer store instruction retired
6988678 r0000000000000800 (47.91%) --> Atomic memory operation retired

1.931335000 seconds time elapsed

1.100415000 seconds user
3.755176000 seconds sys

QEMU:
=========
Perf stat:
=========

[root@fedora-riscv riscv]# perf stat -e r8000000000000005 -e r8000000000000007 \
-e r8000000000000006 -e r0000000000020002 -e r0000000000020004 -e branch-misses \
-e cache-misses -e dTLB-load-misses -e dTLB-store-misses -e iTLB-load-misses \
-e cycles -e instructions perf bench sched messaging -g 15 -l 10 \
Running with 15*40 (== 600) tasks.
Time: 6.578

Performance counter stats for './hackbench -pipe 15 process':

1,794 r8000000000000005 (52.59%) --> SBI_PMU_FW_SET_TIMER
2,859 r8000000000000007 (60.74%) --> SBI_PMU_FW_IPI_RECVD
4,205 r8000000000000006 (68.71%) --> SBI_PMU_FW_IPI_SENT
0 r0000000000020002 (81.69%)
<not counted> r0000000000020004 (0.00%)
<not counted> branch-misses (0.00%)
<not counted> cache-misses (0.00%)
7,878,328 dTLB-load-misses (15.60%)
680,270 dTLB-store-misses (28.45%)
8,287,931 iTLB-load-misses (39.24%)
20,008,506,675 cycles (48.60%)
21,484,427,932 instructions # 1.07 insn per cycle (56.60%)

1.681344735 seconds time elapsed

0.614460000 seconds user
8.313254000 seconds sys

Perf record:
============
[root@fedora-riscv riscv]# perf record -e cycles -e instructions \
-e dTLB-load-misses -e dTLB-store-misses -e iTLB-load-misses -c 10000 \
perf bench sched messaging -g 15 -l 10
# Running 'sched/messaging' benchmark:
# 20 sender and receiver processes per group
# 15 groups == 600 processes run

Total time: 1.261 [sec]
[ perf record: Woken up 1 times to write data ]
[ perf record: Captured and wrote 0.101 MB perf.data (845 samples) ]

[root@fedora-riscv riscv]# perf report
Available samples
407 cycles _
407 instructions _
18 dTLB-load-misses _
2 dTLB-store-misses _
11 iTLB-load-misses _

[1] https://github.com/riscv-non-isa/riscv-sbi-doc/blob/master/riscv-sbi.adoc
[2] https://drive.google.com/file/d/171j4jFjIkKdj5LWcExphq4xG_2sihbfd/edit
[3] https://github.com/atishp04/opensbi/tree/pmu_sscofpmf_v2
[4] https://github.com/atishp04/linux/tree/riscv_pmu_v4
[5] https://github.com/atishp04/qemu/tree/riscv_pmu_v3
[6] https://github.com/atishp04/u-boot/tree/hifive_unmatched_dt_pmu
[7] https://sifive.cdn.prismic.io/sifive/de1491e5-077c-461d-9605-e8a0ce57337d_fu740-c000-manual-v1p3.pdf

Changes from v3->v4:
1. Do not proceed overflow handler if event doesn't set for sampling.
2. overflow status register is only read after counters are stopped.
3. Added the PMU DT node for HiFive Unmatched.

Changes from v2->v3:
1. Added interrupt overflow support.
2. Cleaned up legacy driver initialization.
3. Supports perf record now.
4. Added the DT binding and maintainers file.
5. Changed cpu hotplug notifier to be multi-state.
6. OpenSBI doesn't disable cycle/instret counter during boot. Update the
perf code to disable all the counter during the boot.

Changes from v1->v2
1. Implemented the latest SBI PMU extension specification.
2. The core platform driver was changed to operate as a library while only
sbi based PMU is built as a driver. The legacy one is just a fallback if
SBI PMU extension is not available.

Atish Patra (11):
RISC-V: Remove the current perf implementation
RISC-V: Add CSR encodings for all HPMCOUNTERS
RISC-V: Add a perf core library for pmu drivers
RISC-V: Add a simple platform driver for RISC-V legacy perf
RISC-V: Add RISC-V SBI PMU extension definitions
dt-binding: pmu: Add RISC-V PMU DT bindings
RISC-V: Add perf platform driver based on SBI PMU extension
RISC-V: Add interrupt support for perf
Documentation: riscv: Remove the old documentation
riscv: dts: fu740: Add pmu node
MAINTAINERS: Add entry for RISC-V PMU drivers

.../devicetree/bindings/perf/riscv,pmu.yaml | 51 ++
Documentation/riscv/pmu.rst | 255 ------
MAINTAINERS | 10 +
arch/riscv/Kconfig | 13 -
arch/riscv/boot/dts/sifive/fu740-c000.dtsi | 3 +
arch/riscv/include/asm/csr.h | 66 +-
arch/riscv/include/asm/perf_event.h | 72 --
arch/riscv/include/asm/sbi.h | 97 +++
arch/riscv/kernel/Makefile | 1 -
arch/riscv/kernel/perf_event.c | 485 ------------
drivers/perf/Kconfig | 25 +
drivers/perf/Makefile | 5 +
drivers/perf/riscv_pmu.c | 331 ++++++++
drivers/perf/riscv_pmu_legacy.c | 143 ++++
drivers/perf/riscv_pmu_sbi.c | 732 ++++++++++++++++++
include/linux/cpuhotplug.h | 1 +
include/linux/perf/riscv_pmu.h | 69 ++
17 files changed, 1532 insertions(+), 827 deletions(-)
create mode 100644 Documentation/devicetree/bindings/perf/riscv,pmu.yaml
delete mode 100644 Documentation/riscv/pmu.rst
delete mode 100644 arch/riscv/kernel/perf_event.c
create mode 100644 drivers/perf/riscv_pmu.c
create mode 100644 drivers/perf/riscv_pmu_legacy.c
create mode 100644 drivers/perf/riscv_pmu_sbi.c
create mode 100644 include/linux/perf/riscv_pmu.h

Next message: Masahiro Yamada: "Re: [PATCH 08/10] kbuild: do not include include/config/auto.conf from shell scripts"
Previous message: Atish Patra: "Re: [PATCH 1/2] riscv: prevent null-pointer dereference with sbi_remote_fence_i"
Next in thread: Atish Patra: "Re: [v4 00/11] Improve RISC-V Perf support using SBI PMU and sscofpmf extension"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]