[Bug bootstrap/11932] New: Parallel bootstrap failure with sparc libiberty multilib

[Bug bootstrap/11932] New: Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

           Summary: Parallel bootstrap failure with sparc libiberty multilib
           Product: gcc
           Version: 3.4
            Status: UNCONFIRMED
          Severity: normal
          Priority: P2
         Component: bootstrap
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: lucier at math dot purdue dot edu
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: sparc-sun-solaris2.8
  GCC host triplet: sparc-sun-solaris2.8
GCC target triplet: sparc-sun-solaris2.8

Bootstrap failed with 2003-08-12's gcc sources with make -j 4 on
sparc=sun-solaris2.8 with


Checking multilib configuration...
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libstdc++-v3 ; \
rm -f sparc-sun-solaris2.8/libstdc++-v3/Makefile || : ; \
cp multilib.out sparc-sun-solaris2.8/libstdc++-v3/multilib.out
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libiberty ; \
rm -f sparc-sun-solaris2.8/libiberty/Makefile || : ; \
cp multilib.out sparc-sun-solaris2.8/libiberty/multilib.out
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libf2c ; \
rm -f sparc-sun-solaris2.8/libf2c/Makefile || : ; \
cp multilib.out sparc-sun-solaris2.8/libf2c/multilib.out
mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3
mkdir -p -- sparc-sun-solaris2.8/libiberty
mkdir -p -- sparc-sun-solaris2.8/libf2c
mkdir: No such file or directory
mkdir: Failed to make directory "sparc-sun-solaris2.8/libiberty"; File exists
cp: cannot create sparc-sun-solaris2.8/libiberty/multilib.out: No such file or
directory
make[1]: *** [sparc-sun-solaris2.8/libiberty/multilib.out] Error 2
make[1]: *** Waiting for unfinished jobs....
no
checking how to run the C preprocessor... (cached) gcc -E
checking for ANSI C header files... (cached) yes
checking whether struct tm is in sys/time.h or time.h... time.h
checking for fcntl.h... (cached) yes
checking for unistd.h... (cached) yes
checking for sys/param.h... (cached) yes
checking for stdlib.h... (cached) yes
checking for limits.h... (cached) yes
checking for off_t... (cached) yes
checking whether struct tm is in sys/time.h or time.h... (cached) time.h
checking if mkdir takes one argument... (cached) no
checking size of char... 1
checking size of short... (cached) 2
checking size of int... (cached) 4
checking size of long... (cached) 4
checking size of long long... (cached) 8
checking whether byte ordering is bigendian... (cached) yes
updating cache .././config.cache
creating ./config.status
creating Makefile
creating install-defs.sh
creating config.h
make[1]: Leaving directory
`/export/mdd0/users/lucier/programs/gcc/gcc-3.4/objdir-sparc'
make: *** [bootstrap] Error 2

I restarted a serial bootstrap with 'make bootstrap >>& build.log' that
completed, and I later ran make check.


There seem to be several SMP build problems, and they're not trivial to
diagnose.


Brad

[Bug bootstrap/11932] Parallel bootstrap failure with sparc libiberty multilib

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932


pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|                            |1
           Keywords|                            |build
   Last reconfirmed|0000-00-00 00:00:00         |2003-08-15 19:06:38
               date|                            |


------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-15 19:06 -------
Kaveh R. Ghazi sees the problem too: <http://gcc.gnu.org/ml/gcc/2003-08/msg00898.html>.
Looks like when libstdc++ moved over to autoconf 2.5 needs some variables not be set or 
something.

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[dhazeghi at yahoo dot com]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932


dhazeghi at yahoo dot com changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|normal                      |critical
           Priority|P2                          |P1
            Summary|Parallel bootstrap failure  |[3.4 regression] Parallel
                   |with sparc libiberty        |bootstrap failure with sparc
                   |multilib                    |libiberty multilib


------- Additional Comments From dhazeghi at yahoo dot com  2003-08-23 01:18 -------
Regression. This used to work, not so very long ago.

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[dhazeghi at yahoo dot com]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932



------- Additional Comments From dhazeghi at yahoo dot com  2003-08-24 21:52 -------
Any particular guess when this first started failing? Thanks.

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[pinskia at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932



------- Additional Comments From pinskia at gcc dot gnu dot org  2003-08-25 04:31 -------
This started to happen when libstdc++v3 was converted over to autoconf 2.57.  I think 
that should be reverted untill all libraries have been converted over but that is just my 
opion.

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

PLEASE REPLY TO gcc-bugzilla@gcc.gnu.org ONLY, *NOT* gcc-bugs@gcc.gnu.org.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932



------- Additional Comments From neroden at gcc dot gnu dot org  2003-08-27 21:47 -------
*sigh* No, it shouldn't be reverted; that way lies deadlock.
We should finish converting everything else (note that GCC itself is a *very*
quick conversion, since it currently works 'both ways', and the top level is 
almost as good).  Then we should fix the fallout.

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[steven at gcc dot gnu dot org]

------- Additional Comments From steven at gcc dot gnu dot org  2003-12-16 22:00 -------
Nathane 
> We should finish converting everything else (note that GCC itself 
> is a *very* quick conversion, since it currently works 'both ways', 
> and the top level is almost as good).  Then we should fix the fallout.  
 
Nathanael, OK to assign to you, then?  I saw you posted patches to convert 
gcc/, so this should be fixable soon? 
 
Note that this PR is a P1 regression.  You get double bonus points when you 
can close it!  :-) 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |neroden at gcc dot gnu dot
                   |                            |org


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2003-12-19 10:11 -------
*** Bug 12088 has been marked as a duplicate of this bug. ***

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2003-12-31 21:43 -------
Does it still work without parallel build?  If so, at least there is a 
documented workaround. 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

------- Additional Comments From lucier at math dot purdue dot edu  2003-12-31 21:51 -------
Subject: Re:  [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib


On Dec 31, 2003, at 4:43 PM, neroden at gcc dot gnu dot org wrote:

> Does it still work without parallel build?

yes.



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-01-03 18:23 -------
Subject: Bug 11932

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	neroden@gcc.gnu.org	2004-01-03 18:23:22

Modified files:
	.              : ChangeLog configure.in configure 

Log message:
	PR bootstrap/11932, PR bootstrap/11933
	(I don't know if it will fix either of them, but it relates
	to them.)
	* configure.in: Don't use shared config.cache for target
	directories.
	* configure: Regenerate.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/ChangeLog.diff?cvsroot=gcc&r1=1.845&r2=1.846
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/configure.in.diff?cvsroot=gcc&r1=1.263&r2=1.264
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/configure.diff?cvsroot=gcc&r1=1.123&r2=1.124



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-03 21:40 -------
OK, I found time to actually analyze this.  Looking at the original bug report 
here, there's something very funny going on. 
 
Checking multilib configuration... 
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libstdc++-v3 ; \ 
rm -f sparc-sun-solaris2.8/libstdc++-v3/Makefile || : ; \ 
cp multilib.out sparc-sun-solaris2.8/libstdc++-v3/multilib.out 
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libiberty ; \ 
rm -f sparc-sun-solaris2.8/libiberty/Makefile || : ; \ 
cp multilib.out sparc-sun-solaris2.8/libiberty/multilib.out 
/local/bin/bash ../mkinstalldirs sparc-sun-solaris2.8/libf2c ; \ 
rm -f sparc-sun-solaris2.8/libf2c/Makefile || : ; \ 
cp multilib.out sparc-sun-solaris2.8/libf2c/multilib.out 
^^^^ 
This stuff is the top-level work copying multilib.out, and it's correct. 
 
mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3 
mkdir -p -- sparc-sun-solaris2.8/libiberty 
mkdir -p -- sparc-sun-solaris2.8/libf2c 
^^^^ 
These messages look like they're coming from mkinstalldirs, which is running 
mkdir -p. 
 
mkdir: No such file or directory 
mkdir: Failed to make directory "sparc-sun-solaris2.8/libiberty"; File exists 
^^^^ 
What the heck is up with these?  Obviously the mkdir -p invocations are 
failing.  Can you point me to the Solaris 8 manpage for mkdir? 
 
cp: cannot create sparc-sun-solaris2.8/libiberty/multilib.out: No such file or 
directory 
^^^^ 
This is because sparc-sun-solaris2.8/libiberty didn't get created by 'mkdir 
-p'. 
 
make[1]: *** [sparc-sun-solaris2.8/libiberty/multilib.out] Error 2 
 
 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
   Last reconfirmed|2003-11-26 02:16:10         |2004-01-03 21:43:24
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-05 06:11 -------
OK. Could you try running a fresh sparc-solaris2.8 build with the current 
mainline, and attaching the end of the log? 
I've changed stuff, and I have no idea whether it helps or not, but it should 
at least help with diagnosis.  :-P 
 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[pinskia at gcc dot gnu dot org]

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
           Severity|critical                    |normal


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

------- Additional Comments From lucier at math dot purdue dot edu  2004-01-06 05:21 -------
Subject: Re:  [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib


On Monday, January 5, 2004, at 01:11  AM, neroden at gcc dot gnu dot 
org wrote:

> OK. Could you try running a fresh sparc-solaris2.8 build with the 
> current
> mainline, and attaching the end of the log?

I just did a make -j 12 bootstrap with something checked out just 
before midnight UTC on January 5, i.e., just a few hours ago.  The 
bootstrap did not fail.  (An attempt at a parallel bootstrap yesterday 
did fail.)

So what information would you like from this run?  I'll try a parallel 
bootstrap again later to see if this success was a fluke ;-).

Brad



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-06 23:32 -------
Hah, that's fun.  I guess I might have fixed this with my recent changes
(although that wasn't the immediate intent of the changes).

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-09 01:44 -------
OK, sounds like this is working.  Closing.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

------- Additional Comments From lucier at math dot purdue dot edu  2004-01-13 03:19 -------
Well, it's not working.  I just got exactly the same error.

I'll wait until you tell me what information you need to debug this.

Brad

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-13 23:13 -------
Yaargh. 
 
Not sure what I need here. 
 
Make sure you have a clean checkout and a clean build directory.  :-/  And 
that you're using GNU make.  And a working shell (preferably bash; or ksh). 
 
It seriously looks like a bug in mkdir.  I can't reproduce the "mkdir: No such 
file or directory" or figure out where it's coming from; that should *never*  
happen with mkdir -p. 
Perhaps Solaris 'mkdir -p' doesn't support -- , isn't thread-safe, can't 
handle directories which already exist, or doesn't actually work? 
 
Try running this and see if it works: 
mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3 
If it fails, we have our bug. 
Try it again to make sure it works in all these cases: 
(a) neither directory exists 
(b) sparc-sun-solaris2.8 exists and sparc-sun-solaris2.8/libstdc++-v3 doesn't 
(b) both directories exist 
 
If that works, try running these all at the same time (how best to do this I 
don't know), under the above situations: 
mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3 
mkdir -p -- sparc-sun-solaris2.8/libiberty 
mkdir -p -- sparc-sun-solaris2.8/libf2c 
If one of these fails, this indicates that mkdir -p isn't thread-safe. 
If so, it's not GCC's problem.  :-/ 
 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

------- Additional Comments From lucier at math dot purdue dot edu  2004-01-14 00:58 -------
Subject: Re:  [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

> Try running this and see if it works:
> mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3
> If it fails, we have our bug.
> Try it again to make sure it works in all these cases:
> (a) neither directory exists
> (b) sparc-sun-solaris2.8 exists and sparc-sun-solaris2.8/libstdc++-v3 
> doesn't
> (b) both directories exist

All these work.

> If that works, try running these all at the same time (how best to do 
> this I
> don't know), under the above situations:
> mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3
> mkdir -p -- sparc-sun-solaris2.8/libiberty
> mkdir -p -- sparc-sun-solaris2.8/libf2c

Hmmm.  Don't know how to do this.

I moved my objdir directory where I built this to

http://www.math.purdue.edu/~lucier/objdir

Perhaps the config.cache files or other things (the directories that 
were made?) there can give you some information.

Brad



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at twcny dot rr dot com]

------- Additional Comments From neroden at twcny dot rr dot com  2004-01-14 05:57 -------
Subject: Re:  [3.4 regression] Parallel bootstrap failure
 with sparc libiberty multilib

lucier at math dot purdue dot edu wrote:
> ------- Additional Comments From lucier at math dot purdue dot edu  2004-01-14 00:58 -------
> Subject: Re:  [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib
> 
> 
>>Try running this and see if it works:
>>mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3
>>If it fails, we have our bug.
>>Try it again to make sure it works in all these cases:
>>(a) neither directory exists
>>(b) sparc-sun-solaris2.8 exists and sparc-sun-solaris2.8/libstdc++-v3 
>>doesn't
>>(b) both directories exist
> 
> 
> All these work.
> 
> 
>>If that works, try running these all at the same time (how best to do 
>>this I
>>don't know), under the above situations:
>>mkdir -p -- sparc-sun-solaris2.8/libstdc++-v3
>>mkdir -p -- sparc-sun-solaris2.8/libiberty
>>mkdir -p -- sparc-sun-solaris2.8/libf2c
> 
> 
> Hmmm.  Don't know how to do this.
> 
> I moved my objdir directory where I built this to
> 
> http://www.math.purdue.edu/~lucier/objdir
> 
> Perhaps the config.cache files or other things (the directories that 
> were made?) there can give you some information.
> 
> Brad

Could you try running
mkdir -p -- sparc-sun-solaris2.8/libiberty
when the 'sparc-sun-solaris2.8' directory exists *and* a 'libiberty' 
directory exists at the same level (not below)?  I'm clutching at straws 
now.  The fact that sparc-sun-solaris2.8/libiberty was *not* made is 
clearly the problem, but I can't see for the life of me *why* it wasn't 
made.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-27 06:03 -------
Fair warning: I'm completely at a loss as to what's going on here, and this 
isn't going to get solved until I can diagnose it!  If anyone can help by 
figuring out why mkdir -p is failing, *please* tell me. 

-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

Re: [Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[Ian Lance Taylor]

"neroden at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org> writes:

> ------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-27 06:03 -------
> Fair warning: I'm completely at a loss as to what's going on here, and this 
> isn't going to get solved until I can diagnose it!  If anyone can help by 
> figuring out why mkdir -p is failing, *please* tell me. 

It seems possible that the Solaris mkdir is not robust in the face of
a directory being created after mkdir confirmed that the directory did
not exist.  That would at least explain the ``File exists'' error.

The obvious workaround would be to create $(TARGET_SUBDIR) before
calling mkinstalldirs on $(TARGET_SUBDIR)/DIR.  That is what used to
happen--$(TARGET_SUBDIR) used to be created by configure.  Why don't
you try that and see if it fixes the problem.  Alternatively, try
commenting out the code in mkinstalldirs which execs mkdir -p, and see
if that fixes the problem--the mkinstalldirs shell script appears to
be robust.

Ian

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[ian at wasabisystems dot com]

------- Additional Comments From ian at wasabisystems dot com  2004-01-27 06:22 -------
Subject: Re:  [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

"neroden at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org> writes:

> ------- Additional Comments From neroden at gcc dot gnu dot org  2004-01-27 06:03 -------
> Fair warning: I'm completely at a loss as to what's going on here, and this 
> isn't going to get solved until I can diagnose it!  If anyone can help by 
> figuring out why mkdir -p is failing, *please* tell me. 

It seems possible that the Solaris mkdir is not robust in the face of
a directory being created after mkdir confirmed that the directory did
not exist.  That would at least explain the ``File exists'' error.

The obvious workaround would be to create $(TARGET_SUBDIR) before
calling mkinstalldirs on $(TARGET_SUBDIR)/DIR.  That is what used to
happen--$(TARGET_SUBDIR) used to be created by configure.  Why don't
you try that and see if it fixes the problem.  Alternatively, try
commenting out the code in mkinstalldirs which execs mkdir -p, and see
if that fixes the problem--the mkinstalldirs shell script appears to
be robust.

Ian


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[ghazi at gcc dot gnu dot org]

------- Additional Comments From ghazi at gcc dot gnu dot org  2004-02-04 17:04 -------
I saw this bug again today on the 3.4 branch on sparc-sun-solaris2.7.

FYI, for me it only happens about once per month using -j10 or -j20 
bootstraps.  So it's hard to say whether the mkinstalldirs patch fixes it for 
sure.  I'll have to put the patch in my tree and let a couple of months go by 
without seeing the failure to form an opinion. :-)

Another option would be to somehow serialize the creation of these directories 
so we won't encounter the race condition.




-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |ghazi at caip dot rutgers
                   |                            |dot edu
   Last reconfirmed|2004-01-03 21:43:24         |2004-02-04 17:04:27
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at twcny dot rr dot com]

------- Additional Comments From neroden at twcny dot rr dot com  2004-02-05 00:27 -------
Subject: Re:  [3.4/3.5 regression] Parallel bootstrap
 failure with sparc libiberty multilib

ghazi at gcc dot gnu dot org wrote:
> ------- Additional Comments From ghazi at gcc dot gnu dot org  2004-02-04 17:04 -------
> I saw this bug again today on the 3.4 branch on sparc-sun-solaris2.7.
> 
> FYI, for me it only happens about once per month using -j10 or -j20 
> bootstraps.  So it's hard to say whether the mkinstalldirs patch fixes it for 
> sure.  I'll have to put the patch in my tree and let a couple of months go by 
> without seeing the failure to form an opinion. :-)
> 
> Another option would be to somehow serialize the creation of these directories 
> so we won't encounter the race condition.

That's harder to do correctly than it looks.  But worse, it amounts to 
pretty much the same thing as the mkinstalldirs patch, only done in lots 
of different places ad-hoc.  :-(  Which is why I prefer the 
mkinstalldirs change.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[ghazi at caip dot rutgers dot edu]

------- Additional Comments From ghazi at caip dot rutgers dot edu  2004-02-05 02:21 -------
Subject: Re:  [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

Okay, I'll put the mkinstalldirs patch in my mainline and 3.4 cvs
trees and see if I get the error in the next few months.

Note: there are four copies of mkinstalldirs in the cvs tree!  Your
patch only tweeks the top level one.  I don't know if any of the other
copies are used when this bug triggers or not.  Regardless, you may
want to consolidate these into one copy at the top level seperately
from this patch.  (Like the move-if-change consolidation posted
recently.)



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-02-20 02:01 -------
Subject: Bug 11932

CVSROOT:	/cvs/gcc
Module name:	gcc
Changes by:	neroden@gcc.gnu.org	2004-02-20 02:01:16

Modified files:
	.              : ChangeLog mkinstalldirs install-sh 

Log message:
	PR bootstrap/11932
	* mkinstalldirs, install-sh: Import from automake CVS HEAD.

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/ChangeLog.diff?cvsroot=gcc&r1=1.874&r2=1.875
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/mkinstalldirs.diff?cvsroot=gcc&r1=1.3&r2=1.4
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/install-sh.diff?cvsroot=gcc&r1=1.5&r2=1.6



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-02-20 02:05 -------
A fancy-schmancy fix went upstream, and I grabbed it. 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

[lucier at math dot purdue dot edu]

------- Additional Comments From lucier at math dot purdue dot edu  2004-02-20 02:12 -------
Subject: Re:  [3.4/3.5 regression] Parallel bootstrap failure with sparc libiberty multilib

This bug was reported against 3.4, and the fix seems to have been 
installed only against mainline (?), so I don't think this bug should 
be closed yet.

Brad



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[pinskia at gcc dot gnu dot org]

------- Additional Comments From pinskia at gcc dot gnu dot org  2004-02-20 02:13 -------
Reopening for now.

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |REOPENED
         Resolution|FIXED                       |
            Summary|[3.4/3.5 regression]        |[3.4 regression] Parallel
                   |Parallel bootstrap failure  |bootstrap failure with sparc
                   |with sparc libiberty        |libiberty multilib
                   |multilib                    |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[neroden at gcc dot gnu dot org]

------- Additional Comments From neroden at gcc dot gnu dot org  2004-02-25 23:31 -------
I'm checking in a (slightly) more conservative fix on 3.4. 
 

-- 
           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|REOPENED                    |RESOLVED
         Resolution|                            |FIXED


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932

[Bug bootstrap/11932] [3.4 regression] Parallel bootstrap failure with sparc libiberty multilib

[cvs-commit at gcc dot gnu dot org]

------- Additional Comments From cvs-commit at gcc dot gnu dot org  2004-02-25 23:46 -------
Subject: Bug 11932

CVSROOT:	/cvs/gcc
Module name:	gcc
Branch: 	gcc-3_4-branch
Changes by:	neroden@gcc.gnu.org	2004-02-25 23:46:09

Modified files:
	.              : ChangeLog mkinstalldirs 

Log message:
	PR bootstrap/11932
	* mkinstalldirs: Update to version 2004-02-15.20

Patches:
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/ChangeLog.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.856.2.9&r2=1.856.2.10
http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/mkinstalldirs.diff?cvsroot=gcc&only_with_tag=gcc-3_4-branch&r1=1.3&r2=1.3.12.1



-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11932