Message-ID: <19970630162210.52898@gil.physik.rwth-aachen.de> Date: Mon, 30 Jun 1997 16:22:10 +0200 From: Christoph Kukulies To: Andrew Crabtree Cc: Eric Liao , djgpp AT delorie DOT com Subject: Re: weird DJGPP errors, please help! References: <33b1a6e5 DOT 5673107 AT netnews DOT worldnet DOT att DOT net> <199706301345 DOT AA072748338 AT typhoon DOT rose DOT hp DOT com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii In-Reply-To: <199706301345.AA072748338@typhoon.rose.hp.com>; from Andrew Crabtree on Mon, Jun 30, 1997 at 06:45:38AM -0800 Precedence: bulk I vaguely recall that you get this kind of error when some redirection is active (either something like 2r1 in v1 or some command line redirection to a file). It looks like the as phase is passed some C code. On Mon, Jun 30, 1997 at 06:45:38AM -0800, Andrew Crabtree wrote: > > > > Assembler messages: > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: Ignoring junk 'h' > > after expression > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: operands given don't > > match any known 386 instruction > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: invalid char '[' > > begining second operand '[xoff]' > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: invalid char '[' > > begining second operand '[yoff]' > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: invalid char '[' > > begining second operand '[p]' > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: Ignoring junk 'h' > > after expression > > c:/djgpp/tmp/RHfaaaaa\cccaaaaa(451) Error: Error: operands given don't > > match any known 386 instruction > > Just a guess (post your source next time) > but I'd say you have inline assembly in your program. > DJGPP uses AT&T syntax, which is quite different. See the FAQ for > details on how to convert. > > Off the top of my head - > > Post a b,w,or l after each instruction mnemonic to indicate the size > and do not use word ptr type size overrides. > > registers must be preceded with a %. > Immediates must be preceded with a $ > Source and destination are reversed. > Memory addressing is (I think) > displacement(base,index,scale) > Note that you use parenthesis and commas to separate > > so > > inc word ptr [esi] > > becomes > > incw (%esi) > > HTH > > Andrew > > -- Chris Christoph P. U. Kukulies kuku AT gil DOT physik DOT rwth-aachen DOT de