| --- Log opened Thu Jul 07 00:00:27 2016 | ||
| blueCmd | wallento: did the transfer work? | 01:11 |
|---|---|---|
| olof | Started looking into some semi-completed refactorings of wb_sdram_ctrl and wb_altera_ddr_wrapper that I did two years ago | 02:58 |
| olof | This is a deep deep rabbit hole | 03:04 |
| bandvig | Hello all. Let's return to problem with internal compiler error we discussed here http://juliusbaxter.net/openrisc-irc/%23openrisc.2016-07-02.log.html | 13:39 |
| bandvig | I've tried again to compile NewLIB by modification newlib_cflags with adding -mhard_float. | 13:39 |
| bandvig | I've found that the same error occurs while compiling "/newlib/newlib/libm/math/ef_scalb.c" (In function ‘__ieee754_scalbf’). | 13:39 |
| bandvig | The function declaration "float __ieee754_scalbf(float x, float fn)". | 13:39 |
| bandvig | Additionally, I see, that a lot of files were successfully compiled from the same directories before the error occurs. | 13:40 |
| bandvig | It looks like the problem files (ef_scalb.c, cprojf.c) include "fdlibm.h" while successfully compiled files don't. | 13:40 |
| olofk | Hmm.. bandvig is gone | 17:30 |
| olofk | I made some progress on the hard-float bug | 17:30 |
| olofk | If I'm not here tomorrow, could someone paste him this http://pastie.org/10901488 | 17:32 |
| olofk | That fails. | 17:33 |
| olofk | If I remove the -isystem line, compilation works fine. It also works fine if I remove -mhard-float, or if I pass a constant directly, e.g. isinf(0.0) instead of isinf(f) | 17:34 |
| olofk | Interested in knowing if others can reproduce the bug | 17:35 |
| olofk | Bug is that gcc crashes with an internal error | 17:35 |
| --- Log closed Fri Jul 08 00:00:29 2016 | ||
Generated by irclog2html.py 2.15.2 by Marius Gedminas - find it at mg.pov.lt!