Julian Seward 7f5121f3ce * Fix flag computation for neg instructions ...

* Fix flag and result computation for 16/8-bit shl/shr/sar by amounts
  greater than the word size, which inconveniently is a well-defined
  (if useless) operation on x86.  Sigh.  Unfortunately the solution
  depends on IR shifts behaving as per x86, whereas it would be better
  to define IR shifts only to make sense for values in 0 .. 2^word_size-1.
  This needs to be fixed.

* Implement another shrd case.  shrdl doesn't work correctly 
  according to the QEMU test suite.  I don't know why.



git-svn-id: svn://svn.valgrind.org/vex/trunk@158

2004-07-29 22:26:03 +00:00

VEX

* Fix flag computation for neg instructions

2004-07-29 22:26:03 +00:00

Description

No description provided

51 MiB

Languages

C 94.6%

Assembly 1.7%

C++ 1.1%

Makefile 0.6%

Perl 0.5%

Other 1.4%