[Qemu-devel] [Bug 1248376] Re: "fstenv" gets the wrong result in qemu 1.6.1

[Stefan Weil]

*** This bug is a duplicate of bug 661696 ***
    https://bugs.launchpad.net/bugs/661696

** This bug has been marked a duplicate of bug 661696
   incomplete emulation of fstenv under TCG

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1248376

Title:
  "fstenv" gets the wrong result in qemu 1.6.1

Status in QEMU:
  New

Bug description:
  The float instruction "fstenv" gets the wrong result, when I tested it
  in QEMU 1.6.1.

  The detailed testing environments are: 
  1)the QEMU's guest OS is 32-bit windows 7 professional.(version: 6.1.7600). 
  2)the QEMU's host machine I used is 64-bit Fedora Core 13.

  
  I used the following command line to compile QEMU 1.6.1:
  ./configure --target-list=i386-softmmu
  make 
  make install

  Then, I launched the QEMU guest with the following command:
  qemu-system-i386  xxx.img  -m  1024M


  Here are some more details:
  03FA0005     DAC7                               fcmovb st,st(7)
  03FA0007     D97424 F4                     fstenv (28-byte) ptr ss:[esp-0xC]
  03FA000B     5B                                    pop ebx
  03FA000C     33C9                               xor ecx,ecx

  when 0x03FA000B: pop ebx finished excuting,  the value of ebx should
  be 0x03FA0005. But in QEMU 1.6.1+windows 7 pro(32 bit) which was
  described above, ebx's wrongly turned out to be 0x0!

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1248376/+subscriptions