From: SourceForge.net <no...@so...> - 2003-04-04 05:21:00
|
Bugs item #715085, was opened at 2003-04-04 05:12 You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=100599&aid=715085&group_id=599 Category: pic target Group: None Status: Open Resolution: None Priority: 5 Submitted By: Steve Tell (sgtell) Assigned to: Nobody/Anonymous (nobody) Summary: (PIC) pointers to function breakage; patch included Initial Comment: In the PIC backend, taking a poiner to a function generated bad code and calling a function-pointer didn't generate any code at all. The attached patch fixes a lot of this. A regression test is included. The test doesn't actually run because of other bugs - bogus extra data-memory declarations are produced for the code-space function name symbols _call0 and _call1. (This may be related to or same as 670275). hand-editing ptrfunc.asm after compilation and before assembly demonstrates that the function-pointer code is working. ---------------------------------------------------------------------- >Comment By: Steve Tell (sgtell) Date: 2003-04-04 05:35 Message: Logged In: YES user_id=749390 The attached little perl hack can be used to patch up the ptrfunc.asm regression from the unrelated bugs that prevent it from passing. This is ugly, which is why I didn't hack rules for it into the regression makefile and debated not posting it all. But it does make it possible to test the previous patch before fixing the declaration bug. ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=100599&aid=715085&group_id=599 |