digitalmars.D.bugs - [Issue 840] New: incorrect code generation for asm{fidv ST(1), ST;}
- d-bugmail puremagic.com (23/23) Jan 13 2007 http://d.puremagic.com/issues/show_bug.cgi?id=840
- d-bugmail puremagic.com (13/13) Jan 13 2007 http://d.puremagic.com/issues/show_bug.cgi?id=840
- d-bugmail puremagic.com (28/28) Apr 08 2009 http://d.puremagic.com/issues/show_bug.cgi?id=840
http://d.puremagic.com/issues/show_bug.cgi?id=840
Summary: incorrect code generation for asm{fidv ST(1), ST;}
Product: D
Version: 1.00
Platform: PC
OS/Version: Linux
Status: NEW
Keywords: wrong-code
Severity: normal
Priority: P2
Component: DMD
AssignedTo: bugzilla digitalmars.com
ReportedBy: thomas-dloop kuehne.cn
output:
dc f9 fdivr st(1),st
expected output:
dc f1 fdiv st(1),st
test case:
http://dstress.kuehne.cn/run/a/asm_fdiv_01_D.d
--
Jan 13 2007
d-bugmail puremagic.com http://d.puremagic.com/issues/show_bug.cgi?id=840 other test cases with similar issues are: http://dstress.kuehne.cn/run/a/asm_fdivp_01_A.d http://dstress.kuehne.cn/run/a/asm_fdivp_01_B.d http://dstress.kuehne.cn/run/a/asm_fdivr_01_D.d http://dstress.kuehne.cn/run/a/asm_fdivrp_01_A.d http://dstress.kuehne.cn/run/a/asm_fdivrp_01_B.d http://dstress.kuehne.cn/run/a/asm_fsubp_01_A.d http://dstress.kuehne.cn/run/a/asm_fsubp_01_B.d http://dstress.kuehne.cn/run/a/asm_fsubrp_01_A.d http://dstress.kuehne.cn/run/a/asm_fsubrp_01_B.d --
Jan 13 2007
http://d.puremagic.com/issues/show_bug.cgi?id=840
clugdbug yahoo.com.au changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|NEW |RESOLVED
Resolution| |INVALID
Actually, DMD is correct. Here's the output from the old DOS DEBUG.
-u 0100 0102
13A9:0100 DCF9 FDIV ST(1),ST
13A9:0102 DCF1 FDIVR ST(1),ST
The bug is in majority of *nix assemblers. Here's a link.
===========
http://gcc.gnu.org/ml/gcc-bugs/2003-04/msg01454.html
-----
The SystemV/386 SVR3.2 assembler, and probably all AT&T derived
ix86 Unix assemblers, generate floating point instructions with
reversed source and destination registers in certain cases.
Unfortunately, gcc and possibly many other programs use this
reversed syntax, so we're stuck with it.
eg. `fsub %st(3),%st' results in st = st - st(3) as expected, but
`fsub %st,%st(3)' results in st(3) = st - st(3), rather than
the expected st(3) = st(3) - st
This happens with all the non-commutative arithmetic floating point
operations with two register operands, where the source register is
%st, and destination register is %st(i). See FloatDR below.
----------
--
Apr 08 2009