On Wed, Apr 28, 2010 at 1:24 PM, Ullrich von Bassewitz <uz@musoftware.de> wrote: > Prepending an underscore has been a common behaviour, even for very old C > compilers. There are other ways to mangle the names (appending a $ or > similar), but I don't know a single C compiler that doesn't change the > identifiers when generating assembly. Which makes sense for the reasons > explained by Oliver. I do believe you, but even so, I have never before programmed with a toolset where I've had to use a modified name between asm and high-level code. I know that compilers commit all manner of hideous attrocities on identifier names (esp. for linking), but there are quite a few that have managed to prevent the mangling from having any visibility from the code's perspective. (Or at least, they've hidden it from me. :=] ) Come to think of it, cc65 is the first compiler I've used that actually assembles the intermediate code generated by the compiler as a distinct step (although many could be forced to output an assembly file as well). Note: the first compiler I used was Borland Pascal for DOS, back in 1988. And I had never used unix or linux for anything but checking mail with Pine, until 2 years ago. So (whether or not you cared..) that's where I come from. // Steve ---------------------------------------------------------------------- To unsubscribe from the list send mail to majordomo@musoftware.de with the string "unsubscribe cc65" in the body(!) of the mail.Received on Wed Apr 28 23:26:49 2010
This archive was generated by hypermail 2.1.8 : 2010-04-28 23:26:51 CEST