Re: gcc bug?

2010-10-23 Thread Finn Thain
I get the same results... /Volumes/btc-0.12/gcc-4.4/lib/gcc/m68k-linux-gnu/4.4.5/../../../../m68k-linux-gnu/bin/as -m68020 --pcrel -V -Qy -o gen_uuid.o /var/tmp//cc8hhpkh.s GNU assembler version 2.20.1 (m68k-linux-gnu) using BFD version (GNU Binutils) 2.20.1.20100303 /var/tmp//cc8hhpkh.s: Asse

Re: gcc bug?

2010-10-23 Thread Thorsten Glaser
Finn Thain dixit: >Send me the file and I'll see if I can reproduce it. Sent. The .loc are line numbers in the C file, I see, but I fail to be able to produce a reduced testcase… Anyway, it’s 06:09 localtime, so I should be sleeping for a couple of hours already (but this is fun…) which I’ll do

Re: eglibc done, too

2010-10-23 Thread Thorsten Glaser
Finn Thain dixit: >... but I didn't try to build any fortran frontend, which, if I recall, >was where your build came unstuck (for fidoa). libfortran, IIRC. Maybe it works now, maybe just disable fidoa… >Some of the decisions to disable stuff for m68k were taken a long time ago >under rather d

Re: eglibc done, too

2010-10-23 Thread Finn Thain
On Sun, 24 Oct 2010, Thorsten Glaser wrote: > Finn Thain dixit: > > >This step should probably happen after the multilib/multiarch question > >resolved. I don't really know how to help with that. (As I said, I > >reversed debian's multiarch patch in order to build my > >cross-compilers.) > >

Re: gcc bug?

2010-10-23 Thread Finn Thain
On Sun, 24 Oct 2010, Thorsten Glaser wrote: > Excerpt from gen_uuid.s: Send me the file and I'll see if I can reproduce it. Finn > > My binutils are latest sid ones (2.20.1-15). > > bye, > //mirabilos > -- To UNSUBSCRIBE, email to debian-68k-requ...@lists.debian.org with a subject of "un

Re: eglibc done, too

2010-10-23 Thread Thorsten Glaser
Finn Thain dixit: >This step should probably happen after the multilib/multiarch question >resolved. I don't really know how to help with that. (As I said, I >reversed debian's multiarch patch in order to build my cross-compilers.) Mh. I disabled that for now and doko uploaded that. I might eve

gcc bug?

2010-10-23 Thread Thorsten Glaser
While trying to build: util-linux_2.17.2-3.3.dsc make[4]: Entering directory `/root/ul/util-linux-2.17.2/shlibs/uuid/src' CC clear.lo CC compare.lo

Re: eglibc done, too

2010-10-23 Thread Finn Thain
On Sat, 23 Oct 2010, Thorsten Glaser wrote: [snip] > > I propose that I now (see below) get cowbuilder going with the packages > I have, then recompile everything needed for a buildd, This step should probably happen after the multilib/multiarch question resolved. I don't really know how to h

native compilation, was Re: eglibc done, too

2010-10-23 Thread Finn Thain
On Sat, 23 Oct 2010, Thorsten Glaser wrote: > ...they want to stop using [cross-compilers] after the architecture has > been bootstrapped. I *think* they have been bitten by bugs. It isn't really about bugs. They can be fixed. Native compilation allows a port to claim "self-hosting" status. Whi

Re: eglibc done, too

2010-10-23 Thread Joseph S. Myers
On Sat, 23 Oct 2010, Thorsten Glaser wrote: > Sadly, not by a number of quick greps. It's something I remember from > some time ago, and I *think* I've seen cases in gcc's source where > calls were explicitly not done for cross compilers, but cannot find > them right now. If this has changed in th

eglibc done, too

2010-10-23 Thread Thorsten Glaser
Finn Thain dixit: >> There are code generation differences, disabled optimisation passes, etc. > >I'm not aware of code generation differences. Can you be more specific? Sadly, not by a number of quick greps. It’s something I remember from some time ago, and I *think* I’ve seen cases in gcc’s sou