xed bug and crash on 66 e8 0000

xed bug and crash on 66 e8 0000

imagem de ange4771
ange4771
qui, 21/04/2011 - 05:47

Hi,
I'm using#XED version: [$Id: xed-version.c 2589 2010-12-14 04:20:23Z mjcharne $] I was disassembling my opcode tester (http://code.google.com/p/corkami/wiki/UsermodeTest) and it gave a wrong display result on 66 e8 00 00, mentioning it would call a dword instead of a word. so i turned on verbosity to see what happens (-v 5), then Xed crashed on that specific opcode. I was impressed to see that XED supports so many undocumented opcodes ;) good stuff. sincerely, Ange

RSS Superior
2 posts / 0 new
Último post
Para obter mais informações sobre otimizações de compiladores, consulte Aviso sobre otimizações.
imagem de Mark Charney (Intel)
  • Login para postar comentários
Mark Charney (Intel)
qui, 03/11/2011 - 15:55

Apologies for the delay in responding. I am not familiar with that test. I think XED is behaving properly though. 66 e8 0000 is a near call.... or most of one depending on the mode. The SDM appendix A Opcode table lists it as "Jz" which means that the displacement width depends on the effective operand size, but is at most 32b. The effective operand size is modulated by the 66 prefix in 16/32b modes. Without the 66:in 16b mode, that has a 16b displacement. In 32b mode, it has a 32b displacement (dword) and in 64b mode it has a 32b displacement. With the 66:in 16b mode, it has a 32b displacment. In 32b mode it has a 16b displacment (word). And in 64b mode it always has a 32b displacement. Please let me know if you need more information or can provide more information about your issue.

Superior

Faça login para deixar um comentário.