qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Qemu-devel] a problem about translate tb


From: owen
Subject: [Qemu-devel] a problem about translate tb
Date: Mon, 11 May 2009 00:44:51 +0800

Dear Mr Fabrice:
      I am a graduate student of university in china. I  major in computer.
Recently,I found " When QEMU first encounters a piece of target code, it translate it to host code up to
the next jump or instruction modifying the static CPU state in a way that cannot be deduced at translation
time "  in your paper called QEMU, a Fast and portable Dynamic Translator. So do I think that the size of
some translate block is big. I want to print the tb which contains two or more powerpc code.
      I modifying the function called cpu_exec as list:
         tb = tb_find_fast();
 639                 if(tb->size>4)
 640                   { printf("tb size:%d\n",tb->size);
 641                     exit(0);
 642                   }// in cpu-exec.c

   I let the  ppc emulator reset address is 0x100.  And my  testbench  code is as list:
00000100 <_start>:
     100: 3c 60 00 00  lis r3,0
     104: 60 63 00 00  ori r3,r3,0
     108: 7c 60 01 24  mtmsr r3
     10c: 7c 7b 03 a6  mtsrr1 r3
     110: 3c 60 00 00  lis r3,0
     114: 60 63 00 00  ori r3,r3,0
     118: 7c 70 8b a6  mtspr 560,r3
     11c: 3c 60 00 00  lis r3,0
     120: 60 63 08 00  ori r3,r3,2048
     124: 7c 7e 9b a6  mtspr 638,r3
     128: 3c 20 00 40  lis r1,64
     12c: 60 21 00 00  ori r1,r1,0
     130: 7c 00 02 78  xor r0,r0,r0
     134: 94 01 ff fc  stwu r0,-4(r1)
     138: 94 01 ff fc  stwu r0,-4(r1)
     13c: 3c 60 00 00  lis r3,0
     140: 60 63 00 00  ori r3,r3,0
     144: 48 00 5c 09  bl 5d4c <cpu_init_f>
     148: 3c 40 00 50  lis r2,80
     14c: 7c 56 03 a6  mtdec r2
     150: 3c 60 00 2f  lis r3,47
     154: 60 63 c0 14  ori r3,r3,49172
     158: 3c 80 ff ff  lis r4,-1
     15c: 60 84 ff ff  ori r4,r4,65535
     160: 90 83 00 00  stw r4,0(r3)
     164: 3c 60 00 00  lis r3,0
     168: 60 63 20 8c  ori r3,r3,8332
     16c: 7c 68 03 a6  mtlr r3
     170: 4e 80 00 21  blrl
       .......

        I expect that QEMU would exit  because one tb is more than 4(tb only contains one ppc code).
But I can not find this thing take place in the test. I don't understand why QEMU is not encounters a big tb.
 I wonder if you would tell me the reason by Email.
                            Thank you for your kind consideration of this request.

reply via email to

[Prev in Thread] Current Thread [Next in Thread]