29 Коміти

Автор SHA1 Повідомлення Дата
Venkateshwar Reddy Kandula 997b36f5bc [rocprofiler][navi4] Remove navi4x support on rocprofv2. (#307)
* Remove navi4x support on rocprofv2.

* remove gfx12 from build scripts.

* bug fix.

* address comments.

* update changelog

* Update CHANGELOG.md

* Update CHANGELOG.md

* Update CHANGELOG.md

* address comments

Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>

---------

Co-authored-by: Venkateshwar Reddy Kandula <venkateshwar.kandula1306@gmail.com>
Co-authored-by: Swati Rawat <120587655+SwRaw@users.noreply.github.com>
2025-09-22 03:17:29 -05:00
Giovanni Baraldi d63f98bd00 SWDEV-508485: Adding MFMA F8 metric
Change-Id: I947d2645e8dc1544d198a4e0a02500feee10d89d


[ROCm/rocprofiler commit: f19bed4c8c]
2025-01-09 12:18:02 -06:00
Giovanni Baraldi dc25330772 Revert "SWDEV-310289: Adding SPI pipe selection"
This reverts commit 896404efb3.

Reason for revert: Requires priv_cp_queues=1

Change-Id: Ia6c78ac25b88d7ef4703654075d54e672a6e320c


[ROCm/rocprofiler commit: a6328a1481]
2024-10-25 02:57:52 -04:00
Giovanni LB 896404efb3 SWDEV-310289: Adding SPI pipe selection
Change-Id: I4856d284df3dccaa100a2341211ae09e11c63ecd


[ROCm/rocprofiler commit: e5e2c6041d]
2024-10-12 01:14:49 -04:00
Lang Yu 2230af4b1d SWDEV-467545 - Add rocprofiler support for gfx1150/gfx1151
Change-Id: I2cddc36981f6d815c865d180a1daf1b8a7e0633f
Signed-off-by: Lang Yu <lang.yu@amd.com>


[ROCm/rocprofiler commit: 7313e52f35]
2024-07-09 22:40:10 -04:00
jatang c8d58d1986 SWDEV-458392 - Add gfx12 support.
Change-Id: I91bb6a3329bf77f26005a345c18b63b86922028a


[ROCm/rocprofiler commit: e7b96b1e71]
2024-06-17 13:24:48 -04:00
Mythreya 6a4321c4e3 Add MI200/MI300 counters
Revision - Addition [Impact SoC: MI200, MI300]
    Note: this set of counters are important help understand the
          bottleneck.
    1.	TCC_TAG_STALL
        a.	Metric:  TCC_TAG_STALL/TCC_CYCLE: percentage of time TCC
            tag lookup pipeline is stalled
    2.	TCP_TCR_TCP_STALL_CYCLES
        a.	Metric: TCP_TCR_TCP_STALL_CYCLES/TCP_GATE_EN1: percentage
            of time TCP is stalled by TCR

Revision - Addition [Impact SoC: MI300]
    3.	TCC_BUBBLE:
        a.	Definition: Number of 128-byte read requests sent to EA
        b.	Revised Metric #1, TCC-EA Read BW:
        ReadBW =   128 * TCC_BUBBLE
                + 64 * (TCC_EA0_RDREQ - TCC_BUBBLE - TCC_EA0_RDREQ_32B)
                + 32 * TCC_EA0_RDREQ_32B
        c.	Revised Metric #2: TCC_EA Read Latency
        ReadLatency = TCC_EA0_RDREQ_LEVEL / (TCC_BUBBLE + TCC_EA0_RDREQ)

    /* [Fineprint] More detailed arithmetic:
    * ReadLatency = TCC_EA0_RDREQ_LEVEL / (#32B_req + #64B_req + #128B_req * 2)
    */

Change-Id: I0a2dfc1b64ca97023b1e8ba0f9830330b3034946


[ROCm/rocprofiler commit: 46e02a9866]
2023-10-30 15:38:46 -04:00
Mythreya 1471dc8a76 Remove non-functional counters for MI200 and MI300
Counters removed for MI300 (gfx940)
    TCP_TCC_WRITE_REQ_HOLE_LATENCY
    TCP_TCC_WRITE_REQ_LATENCY
    TCP_TCC_READ_REQ_LATENCY
    TCP_TCP_LATENCY

Counters removed for MI200 and MI300 (gfx90a and gfx940 respectively)
    TA_BUFFER_COALESCABLE_WAVEFRONT
    TA_FLAT_COALESCABLE_WAVEFRONT
    TCC_EA0_WRREQ_IO_CREDIT_STALL
    TCC_EA0_WRREQ_GMI_CREDIT_STALL
    TCC_EA0_WRREQ_DRAM_CREDIT_STALL
    TCC_EA0_RDREQ_IO_CREDIT_STALL
    TCC_EA0_RDREQ_GMI_CREDIT_STALL
    TCC_EA0_RDREQ_DRAM_CREDIT_STALL

Change-Id: Ic3d1e7bf35495f35b1239f03ca6420e949421386


[ROCm/rocprofiler commit: 1fae494b12]
2023-10-26 12:50:57 -04:00
Giovanni LB 552814f227 SWDEV-419944: Added metrics for gfx1102
Change-Id: I5c69ff716f530d130710c0687f20e5bc990a60eb


[ROCm/rocprofiler commit: 43e259e5da]
2023-09-11 13:59:33 -04:00
Giovanni LB dda0379742 SWDEV-298742: Added occupancy metrics
Change-Id: I67e375ad06535bbb8cc864b78840ce3962bcc58e


[ROCm/rocprofiler commit: a1508035dc]
2023-06-19 12:10:22 -04:00
Giovanni LB a7e8182a21 SWDEV-405575: Added gfx941 and gfx942
Change-Id: I45a49cd64a76d3ae32c209497c70fe27b5be212b


[ROCm/rocprofiler commit: e1285e3fd4]
2023-06-19 11:11:37 -04:00
Saurabh Verma dcd5f1a397 MI300 counters support for rocprof and rocprofv2(Accumulation from all xccs)
1. Xml files updated for gfx940 counters
2. File plugin changes to allow rocprofv2 backward compatibility for results.csv
3. Changes in rocprofv2 script to use tblextr.py, to generate results.csv just like rocprof

Change-Id: I7798f4411ce01f6fbfffb126de654ed806ca7045
(cherry picked from commit 86cbaf38c436be876f0426fa27803b1e64d90378)


[ROCm/rocprofiler commit: 8f82ff6a46]
2023-05-30 21:41:54 -05:00
Kiumars Sabeti a260b63b96 SWDEV-380635: adding gfx11 architecture to rocprofiler which includes navi31 and navi32 for now
Change-Id: Ib2a93a34688471c82b5db0dc10e8da58452dba21


[ROCm/rocprofiler commit: 997c771723]
2023-05-05 15:39:18 -04:00
Kiumars Sabeti f893fecf87 SWDEV-369023: Added two new counters SQ_INSTS_TEX_LOAD and SQ_INSTS_TEX_STORE for gfx10.These two new counters are replacement for SQ_INSTS_VMEM_RD and SQ_INSTS_VMEM_WR which are not supported in gfx10 architecture
Change-Id: I4c4101eea27f9073492ae42c70a30a002f4d8834


[ROCm/rocprofiler commit: a9a82ee107]
2022-12-09 20:41:45 -05:00
Kiumars Sabeti d5974aba78 SWDEV-302380: [ROCm QA][Mainline][Navi21] 6 tests are failing in rocprofiler-stg2
This is an attempt to support basic and derived counters for navi21.  This code will not work correctly unless we add navi counters to metrics.xml and gfx_metrics.xml

Change-Id: Ied06a81345a6fbb02fa0fde1889d94bbe64e9a03


[ROCm/rocprofiler commit: b53fd84ade]
2022-08-05 17:31:37 -04:00
Saurabh Verma eaebfe0954 SWDEV-297195: Corrected units for some counters. Units changed to quad-cycles units where required.
Change-Id: Ia6b0387ac6ec4210bb9482d85ae5635fc7c3c9d0


[ROCm/rocprofiler commit: 18dedbaee8]
2022-07-21 17:22:17 -05:00
Saurabh Verma e27d5da8c0 SWDEV-298750:Approval to make internal profile counters public
Added approved HW counters for MI200. Also added derived metrics for the same

Change-Id: I1c6abfdfde4e4fd4ba8bd5eec0557ad08fd71c77


[ROCm/rocprofiler commit: 6d233c65d7]
2022-05-17 16:44:16 -05:00
Chun Yang b9d8cc066a SWDEV-324379 : Expose FP64 and FP32 performance counters on on AMD profilers for MI200
Change-Id: I2c38ccc297872dfc1896314ceadbed98dc761766


[ROCm/rocprofiler commit: 26c479c72a]
2022-03-17 14:06:24 -07:00
AMD 9e422660cd Add support for gfx90a
Merge gfx90a support from the 'amd-npi' branch.

Change-Id: I9b51711ed4a1d2f1ed42ba9b83cb12136be228b8


[ROCm/rocprofiler commit: 4df3e0bd9a]
2021-06-16 16:35:42 -07:00
Evgeny 7bf4ab543d unordered metrics expressions config
Change-Id: I998423046abfd7af45f28071911276cce1c4f3da


[ROCm/rocprofiler commit: e5b938b5fb]
2019-08-25 05:08:55 -05:00
Evgeny fc6e208b50 overwriting on inheritance
Change-Id: I6faf79874fc14e898c75972aa2f372e228105f7b


[ROCm/rocprofiler commit: 82ecf9e7a4]
2019-08-08 01:54:29 -05:00
Evgeny 55680ff541 gfx908 events update
Change-Id: Ieb7c72f66b7e5e5dfc29d97650a29722a6276a31


[ROCm/rocprofiler commit: 09eed4f35b]
2019-08-07 19:44:31 -05:00
Xianwei Zhang 62b33801c4 xml: add extra events for gfx906 GPUs
Compard to gfx900, e.g., Vega 10, gfx906 adds extra counter events.
A typical difference is on TCC-EA that gfx906 (e.g., Vega 20) has 2
EAs per TCC, while only one single EA/TCC on gfx900. As such,
additional counters must be profiled to get correct results. This
patch adds one extra events to specifically handle gfx906.

Change-Id: Id6c9d37548a102c80bbfddcfa11e77d20f17431a


[ROCm/rocprofiler commit: ca9a714b77]
2019-07-16 12:56:45 -05:00
Evgeny c3c9815c49 minor changes
Change-Id: I16444de8c5d83a9f33392ac870d2fc188978ab54


[ROCm/rocprofiler commit: ea0a1110dc]
2018-12-18 09:51:47 -06:00
Evgeny cbdf1990eb - cleanup;
- metrics descriptions;

Change-Id: I9087f41a4ba3ed6c94039ac921407aa13529bf6c


[ROCm/rocprofiler commit: 3b7b958a5a]
2018-08-20 00:40:10 -05:00
Evgeny 8507de8fad adding run scripts
Change-Id: Ia7048a0baaa6ea03283aade53a0758d147dc6fcd


[ROCm/rocprofiler commit: deb3b4df0d]
2018-07-01 15:31:42 -05:00
Evgeny e4f873df29 cleanup
Change-Id: Iadb1603a49bfb1b4c3bb1d31cf3ad5f736dda40a


[ROCm/rocprofiler commit: 46d0368d9e]
2018-06-15 12:05:47 -05:00
Evgeny 7ae981cdd2 aqlprofile/rocprofiler util sync
Change-Id: If029a5aa4eff9ad3e01dc83b41426c079f7f03a7


[ROCm/rocprofiler commit: cca3ca68c1]
2018-05-16 13:04:07 -05:00
Evgeny 20d448818e test structure refactoring, unifying with aqlprofile
Change-Id: Icad642413c7edb27885effc9da3524cf3899f8de


[ROCm/rocprofiler commit: 71044909ce]
2018-05-03 16:47:18 -05:00