All of lore.kernel.org
 help / color / mirror / Atom feed
* [parisc-linux] what i did on my holidays
@ 2001-09-27 15:10 Matthew Wilcox
  2001-09-28  0:44 ` Carlos O'Donell Jr.
  0 siblings, 1 reply; 5+ messages in thread
From: Matthew Wilcox @ 2001-09-27 15:10 UTC (permalink / raw)
  To: parisc-linux

i've been hacking on various random things; i don't have sufficient time
to commit everything properly (so i'll do it when i get back on monday),
but the curious can look at `all.diff2' in my homedir on pehc.  here's
a rough contents of the diff (in no particular order):

 * changes in support of card-mode dino
 * random spelling mstakes, documentation updates, etc
 * misc cleanups
 * convert ccio-rm-dma to the parisc_device interface.  it was annoying me.
 * change a lot of `void *' to `unsigned long' and remove casts.
 * rationalise function names & user messages

And much much more!

I haven't even tried compiling it since fixing up the cvs merge conflicts
with all the good work done in the last week or so.  Have fun...

-- 
Revolutions do not require corporate support.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [parisc-linux] what i did on my holidays
  2001-09-27 15:10 [parisc-linux] what i did on my holidays Matthew Wilcox
@ 2001-09-28  0:44 ` Carlos O'Donell Jr.
  2001-09-28 14:23   ` Matthew Wilcox
  0 siblings, 1 reply; 5+ messages in thread
From: Carlos O'Donell Jr. @ 2001-09-28  0:44 UTC (permalink / raw)
  To: parisc-linux

> 
> And much much more!
> 
> I haven't even tried compiling it since fixing up the cvs merge conflicts
> with all the good work done in the last week or so.  Have fun...
> 
> -- 

You da best! :)

I've been hacking glibc 2.2.4-1:

	- ./sysdeps/ia64/dl-fptr.c changes for try_lock
	- Put change back into dpatch
	
	- Combing through glibc trying to understand the various
	  changes that have been made (and how glibc works).

	- dpkg-buildpackage now makes it to the point of using a new
	  ld.so.1, which seg's... so I'm back to figuring out why.

I'm still not as far as you've gotten. I can't even get within arms
reach of running zic, though I'm getting closer.

Trying to document those g++ problems with #define _BSD_SOURCE that
Zimmerman has been trying to assign blame for ;)

I constantly mull over writing a new "building the toolchain" for
the website.

Then I listen to Tom Waits, and I just feel like drinking and pondering
if my 715/50 really loves me ;)

--


c.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [parisc-linux] what i did on my holidays
  2001-09-28  0:44 ` Carlos O'Donell Jr.
@ 2001-09-28 14:23   ` Matthew Wilcox
  2001-09-28 17:37     ` Carlos O'Donell Jr.
  0 siblings, 1 reply; 5+ messages in thread
From: Matthew Wilcox @ 2001-09-28 14:23 UTC (permalink / raw)
  To: Carlos O'Donell Jr., parisc-linux

On Thu, Sep 27, 2001 at 08:44:51PM -0400, Carlos O'Donell Jr. wrote:
> 	- ./sysdeps/ia64/dl-fptr.c changes for try_lock
> 	- Put change back into dpatch

Excellent.

> 	- dpkg-buildpackage now makes it to the point of using a new
> 	  ld.so.1, which seg's... so I'm back to figuring out why.

Yes.  If you decode that long command, it's probably running zic -- that's
the first thing it attempts to build natively.

By the way, there's glibc 2.2.4-2 out now; I don't know what changes
BenC made to it.  In any case, I thought we'd decided the problem was
with a bogus C compiler (since 2.2.3-9 now fails to build).  Any update
on getting a compiler which fixes this?

-- 
Revolutions do not require corporate support.

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [parisc-linux] what i did on my holidays
  2001-09-28 14:23   ` Matthew Wilcox
@ 2001-09-28 17:37     ` Carlos O'Donell Jr.
  2001-09-28 20:11       ` Matt Taggart
  0 siblings, 1 reply; 5+ messages in thread
From: Carlos O'Donell Jr. @ 2001-09-28 17:37 UTC (permalink / raw)
  To: Matthew Wilcox; +Cc: parisc-linux

> > 	- ./sysdeps/ia64/dl-fptr.c changes for try_lock
> > 	- Put change back into dpatch
> 
> Excellent.
> 
> > 	- dpkg-buildpackage now makes it to the point of using a new
> > 	  ld.so.1, which seg's... so I'm back to figuring out why.
> 
> Yes.  If you decode that long command, it's probably running zic -- that's
> the first thing it attempts to build natively.
> 
> By the way, there's glibc 2.2.4-2 out now; I don't know what changes
> BenC made to it.  In any case, I thought we'd decided the problem was
> with a bogus C compiler (since 2.2.3-9 now fails to build).  Any update
> on getting a compiler which fixes this?
>

No, I never get as far as making zic.

make[2]: Entering directory `/raid/glibc_build2/glibc-2.2.4/glibc-2.2.4/sunrpc'
/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/elf/ld.so.1 --library-path /raid/glibc_build2/glibc-2.2.4/hppa-linux/obj:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/math:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/elf:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/dlfcn:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/nss:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/nis:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/rt:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/resolv:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/crypt:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/linuxthreads /raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/sunrpc/rpcgen -Y `gcc -print-file-name=cpp | sed 's|/cpp$||'` -c rpcsvc/bootparam_prot.x -o /raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/sunrpc/xbootparam_prot.T
make[2]: *** [/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/sunrpc/xbootparam_prot.stmp] Segmentation fault
make[2]: Leaving directory `/raid/glibc_build2/glibc-2.2.4/glibc-2.2.4/sunrpc'
make[1]: *** [sunrpc/others] Error 2
make[1]: Leaving directory `/raid/glibc_build2/glibc-2.2.4/glibc-2.2.4'
make: *** [all] Error 2

Is the last I see. And I'm pretty confused by _what_ is's really trying to do.
It's calling ld.so.1? Or is this just really ld? I wigged out the first time
I saw it fail here. I've seen many weird things, but the internals of glibc are
_really_ wired.

execve("/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/elf/ld.so.1", ["/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/elf/ld.so.1", "--library-path", "/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/math:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/elf:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/dlfcn:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/nss:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/nis:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/rt:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/resolv:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/crypt:/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/linuxthreads", "/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/sunrpc/rpcgen", "-Y", "/lib", "-c", "rpcsvc/bootparam_prot.x", "-o", "/raid/glibc_build2/glibc-2.2.4/hppa-linux/obj/sunrpc/xbootparam_prot.T"], [/* 18 vars */]) = 0
--- SIGSEGV (Segmentation fault) ---
+++ killed by SIGSEGV +++

Out of curiosity, is there a gdb package built for debian-hppa?
(Last I checked there was none)

I'll try to debug this on the weekend.
Though my 712/60 has just gone dead on serial and ssh :}

---

I'll give glibc-2.2.4-2 a shot, and see what the difference are.
I assume that -2 is on the debian pool already.

Who is BenC? :)
Is he making general glibc changes or hppa specific ones?

Who's working on the compiler for parisc?

c. 

^ permalink raw reply	[flat|nested] 5+ messages in thread

* Re: [parisc-linux] what i did on my holidays
  2001-09-28 17:37     ` Carlos O'Donell Jr.
@ 2001-09-28 20:11       ` Matt Taggart
  0 siblings, 0 replies; 5+ messages in thread
From: Matt Taggart @ 2001-09-28 20:11 UTC (permalink / raw)
  To: Carlos O'Donell Jr., Matthew Wilcox, parisc-linux

"Carlos O'Donell Jr." writes...

[snip]
> Out of curiosity, is there a gdb package built for debian-hppa?

There's an unofficial package in

ftp://puffin.external.hp.com/pub/parisc/binaries/debian/unofficial-debs

pehc cvs needs to be synced up with upstream before the changes can be sent 
upstream and also put in the official Debian package.

[snip]
> Who's working on the compiler for parisc?

No one. Alan Modra periodically does cvs merges and I use that to generate 
patches for the Debian package. We need to catch up with the 
3.0.2ds2-0pre010922 version of the package, which I hope to do next week. That 
might help glibc but I'm betting whatever's causing the build failure it still 
out there.

-- 
Matt Taggart        Linux Development Lab
taggart@fc.hp.com   HP Linux Systems Operation

^ permalink raw reply	[flat|nested] 5+ messages in thread

end of thread, other threads:[~2001-09-28 20:11 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2001-09-27 15:10 [parisc-linux] what i did on my holidays Matthew Wilcox
2001-09-28  0:44 ` Carlos O'Donell Jr.
2001-09-28 14:23   ` Matthew Wilcox
2001-09-28 17:37     ` Carlos O'Donell Jr.
2001-09-28 20:11       ` Matt Taggart

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.