Mailing List Archives

Authenticated access

 		[Computer Systems Lab]

[← Prev]  [Date]  [Next →] [← Prev]  [Thread]  [Next →]

[Gems-users] Some problems on the trace of MESI_CMP_filter_Directory protocol(fully)

Date: Thu, 10 Dec 2009 16:16:00 +0800
From: ruzhu kao <kaoruzhu@xxxxxxxxx>
Subject: [Gems-users] Some problems on the trace of MESI_CMP_filter_Directory protocol(fully)
Hi all:
   I am sorry to sent the incomplete mail by my misoperation some minutes ago.      
I want to understand the mechanism in MESI_CMP_filter_Directory protocol, I used tester.exec to replay the simple trace like wiki in WISC.
The trace file only contains these instructions below:
  7  0x400 0 LD
  1  0x400 0 LD
  8  0x400 0 LD
12  0x400 0 ST
  1  0x400 0 LD

Then  I use a 4*4 mesh topology for interconnect and adding some configurations in tester.defaults to make it support XACT.
This is partial output from stats file.

      1   7  -1        Seq               Begin       >       [0x400, line 0x400] LD
      2   0   7    L1Cache                Load     NP>IS     [0x400, line 0x400]  request: Timestamp: 0 PA: [0x400, line 0x400] Type: LD VPC: [0x0, line 0x0] Mode: UserMode PF: Yes VA: [0x0, line 0x0] Thread: 0 Exposed: 00
      9   0   0    L2Cache             L1_GETS     NP>ISS    [0x400, line 0x400]  request: Timestamp: 0 Requestor: L1Cache-7 Dest: [NetDest (3) 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  - 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  - 0 0 0 0  - ] PA: [0x400, line 0x400] Type: GETS Mode: UserMode PF: Yes requestor: L1Cache-7 [0x400, line 0x400]
     17   0   0    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     20   0   1    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     23   0  15    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     23   0   3    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     23   0   6    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     23   0   8    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     23   0   2    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     25   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: 0 [0x400, line 0x400] 1 after pendingAcks: -1 sender: L1Cache-0
     26   0   5    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     26   0  14    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     26   0   4    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     26   0   9    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     29   0  11    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     29   0  10    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     29   0  12    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     29   0  13    L1CacheCheck_Read_Write_Filter     NP>NP     [0x400, line 0x400]  [0x400, line 0x400]
     31   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -1 [0x400, line 0x400] 1 after pendingAcks: -2 sender: L1Cache-6
     31   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -2 [0x400, line 0x400] 1 after pendingAcks: -3 sender: L1Cache-1
     31   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -3 [0x400, line 0x400] 1 after pendingAcks: -4 sender: L1Cache-8
     34   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -4 [0x400, line 0x400] 1 after pendingAcks: -5 sender: L1Cache-15
     37   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -5 [0x400, line 0x400] 1 after pendingAcks: -6 sender: L1Cache-9
     37   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -6 [0x400, line 0x400] 1 after pendingAcks: -7 sender: L1Cache-3
     37   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -7 [0x400, line 0x400] 1 after pendingAcks: -8 sender: L1Cache-2
     37   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -8 [0x400, line 0x400] 1 after pendingAcks: -9 sender: L1Cache-5
     40   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -9 [0x400, line 0x400] 1 after pendingAcks: -10 sender: L1Cache-4
     40   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -10 [0x400, line 0x400] 1 after pendingAcks: -11 sender: L1Cache-14
     43   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -11 [0x400, line 0x400] 1 after pendingAcks: -12 sender: L1Cache-10
     46   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -12 [0x400, line 0x400] 1 after pendingAcks: -13 sender: L1Cache-13
     46   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -13 [0x400, line 0x400] 1 after pendingAcks: -14 sender: L1Cache-11
     46   0   7    L1Cache                 Ack     IS>IS     [0x400, line 0x400]  before pendingAcks: -14 [0x400, line 0x400] 1 after pendingAcks: -15 sender: L1Cache-12
     53   0   0  Directory               Fetch      I>I      [0x400, line 0x400] 
    210   0   0    L2Cache            Mem_Data    ISS>ISS_MB [0x400, line 0x400] 
    230   0   7    L1Cache   L2_Exclusive_Data     IS>IS_E   [0x400, line 0x400]  before pendingAcks: -15 [0x400, line 0x400] -15 after pendingAcks: 0 sender: L2Cache-0 Triggering All_Acks
    231   7  -1        Seq                Done       >       [0x400, line 0x400] 230 cycles NULL LD Yes
    231   0   7    L1Cache             Ack_all   IS_E>E      [0x400, line 0x400] 
    237   0   0    L2Cache   Exclusive_Unblock ISS_MB>MT     [0x400, line 0x400] 
    301   1  -1        Seq               Begin       >       [0x400, line 0x400] LD
    302   0   1    L1Cache                Load     NP>IS     [0x400, line 0x400]  request: Timestamp: 0 PA: [0x400, line 0x400] Type: LD VPC: [0x0, line 0x0] Mode: UserMode PF: Yes VA: [0x0, line 0x0] Thread: 0 Exposed: 00
    309   0   0    L2Cache             L1_GETS     MT>MT_IIB [0x400, line 0x400]  request: Timestamp: 0 Requestor: L1Cache-1 Dest: [NetDest (3) 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  - 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0  - 0 0 0 0  - ] PA: [0x400, line 0x400] Type: GETS Mode: UserMode PF: Yes requestor: L1Cache-1 [0x400, line 0x400]
    320   0   7    L1Cache            Fwd_GETS      E>S      [0x400, line 0x400] 
    328   0   0    L2Cache       WB_Data_clean MT_IIB>MT_SB  [0x400, line 0x400] 
    331   1  -1        Seq                Done       >       [0x400, line 0x400] 30 cycles NULL LD Yes 
 
My question is why the first LD 0x400 on Processor 7 need to run L1CacheCheck_Read_Write_Filter  on other processors. After the first LD on Processor 7 L2Cache transform from ISS_MB to MT, which indicates "L2 cache entry Modified in a local L1, assume L2 copy stale" It make me very puzzle, I used to think LD just need to run L1CacheCheck_Write_Filter and L2 cache should be in SS.
Can anybody explain it why does it happened?

Another problem is when I look through the protocol file, I find a problem on access permission in *L1cache.sm which *E* state is set to be read-only. *E* state in MESI protocol should be Read-writable, and I do not know the reason changing it here.

I am looking forward to heard any explanations on these problems. Thanks in advance.


Sincerely yours:
Ruzhu
[← Prev in Thread] Current Thread [Next in Thread→]
  • [Gems-users] Some problems on the trace of MESI_CMP_filter_Directory protocol(fully), ruzhu kao <=
Previous by Date:  Re: [Gems-users] Multi-processor cache statistics, Philip Garcia
Next by Date:  Re: [Gems-users] Multi-processor cache statistics, Vishal Oliyil Kunnil
Previous by Thread:  [Gems-users] Some problems on the trace of MESI_CMP_filter_Directory protocol, ruzhu kao
Next by Thread:  [Gems-users] TSO deprecated?, Rakesh Komuravelli
Indexes:  [Date] [Thread]