public inbox for overseers@sourceware.org
 help / color / mirror / Atom feed
* cvs cache corrupt again
@ 2001-11-09 10:38 Richard Henderson
  2001-11-09 10:47 ` Christopher Faylor
                   ` (2 more replies)
  0 siblings, 3 replies; 12+ messages in thread
From: Richard Henderson @ 2001-11-09 10:38 UTC (permalink / raw)
  To: overseers

cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gcc/ChangeLog,v'


r~

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 10:38 cvs cache corrupt again Richard Henderson
@ 2001-11-09 10:47 ` Christopher Faylor
  2001-12-17  9:27   ` Christopher Faylor
  2001-11-09 11:29 ` Angela Marie Thomas
  2001-12-17  9:19 ` Richard Henderson
  2 siblings, 1 reply; 12+ messages in thread
From: Christopher Faylor @ 2001-11-09 10:47 UTC (permalink / raw)
  To: Richard Henderson; +Cc: overseers

On Mon, Dec 17, 2001 at 09:18:22AM -0800, Richard Henderson wrote:
>cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gcc/ChangeLog,v'

I nuked fileattr.  It seems to be ok now.

cgf

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 10:38 cvs cache corrupt again Richard Henderson
  2001-11-09 10:47 ` Christopher Faylor
@ 2001-11-09 11:29 ` Angela Marie Thomas
  2001-11-09 12:03   ` Jason Molenda
                     ` (2 more replies)
  2001-12-17  9:19 ` Richard Henderson
  2 siblings, 3 replies; 12+ messages in thread
From: Angela Marie Thomas @ 2001-11-09 11:29 UTC (permalink / raw)
  To: Richard Henderson; +Cc: overseers

> cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
c/ChangeLog,v'

I was told once that this happened when non-fileattr-aware versions of
cvs were used to make commits.  If that's the case, this will happen a
lot (which it seems to).  I can't think of any reason why the client would
have to know about it, so maybe it's a bug?

Chris, when you were looking at the local hacks, did you notice anything
that would indicate the client needed support for fileattr as well?

--Angela

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 11:29 ` Angela Marie Thomas
@ 2001-11-09 12:03   ` Jason Molenda
  2001-12-17  9:49     ` Jason Molenda
  2001-11-09 12:49   ` Ian Lance Taylor
  2001-12-17  9:39   ` Angela Marie Thomas
  2 siblings, 1 reply; 12+ messages in thread
From: Jason Molenda @ 2001-11-09 12:03 UTC (permalink / raw)
  To: angela; +Cc: Richard Henderson, overseers

On Mon, Dec 17, 2001 at 09:37:40AM -0800, Angela Marie Thomas wrote:
> > cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
> c/ChangeLog,v'
> 
> I was told once that this happened when non-fileattr-aware versions of
> cvs were used to make commits.  

It's only the server that matters, and we've carefully hidden away
all unpatched cvs binaries on sourceware.  It's probably just a
bug - one of these incorrect fileattrs comes up every couple of
months.  The cvs patch could be rewritten to unlink the fileattr
cache if there is a mismatch like this.  That wouldn't be appropriate
at most sites (some other cvs server-side state can be stored in
the fileattr files), but it'd be harmless on sourceware.

Jason

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 11:29 ` Angela Marie Thomas
  2001-11-09 12:03   ` Jason Molenda
@ 2001-11-09 12:49   ` Ian Lance Taylor
  2001-11-09 14:57     ` Phil Edwards
  2001-12-17 10:02     ` Ian Lance Taylor
  2001-12-17  9:39   ` Angela Marie Thomas
  2 siblings, 2 replies; 12+ messages in thread
From: Ian Lance Taylor @ 2001-11-09 12:49 UTC (permalink / raw)
  To: angela; +Cc: overseers

Angela Marie Thomas <angela@dragaera.ReleaseDominatrix.com> writes:

> > cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
> c/ChangeLog,v'
> 
> I was told once that this happened when non-fileattr-aware versions of
> cvs were used to make commits.  If that's the case, this will happen a
> lot (which it seems to).  I can't think of any reason why the client would
> have to know about it, so maybe it's a bug?
> 
> Chris, when you were looking at the local hacks, did you notice anything
> that would indicate the client needed support for fileattr as well?

The client doesn't need to know about the fileattrs patch.  Only the
CVS which is modifying the actual repository.

This problem can happen if a commit is interrupted at the wrong time,
or more generally if the commit fails for some reason at the wrong
time.

Ian

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 12:49   ` Ian Lance Taylor
@ 2001-11-09 14:57     ` Phil Edwards
  2001-12-17 10:56       ` Phil Edwards
  2001-12-17 10:02     ` Ian Lance Taylor
  1 sibling, 1 reply; 12+ messages in thread
From: Phil Edwards @ 2001-11-09 14:57 UTC (permalink / raw)
  To: Ian Lance Taylor; +Cc: angela, overseers

On Mon, Dec 17, 2001 at 10:01:50AM -0800, Ian Lance Taylor wrote:
> This problem can happen if a commit is interrupted at the wrong time,
> or more generally if the commit fails for some reason at the wrong
> time.

Yeah, somebody was talking about a modem dying mid-commit, I believe.

-- 
If ye love wealth greater than liberty, the tranquility of servitude greater
than the animating contest for freedom, go home and leave us in peace.  We seek
not your counsel, nor your arms.  Crouch down and lick the hand that feeds you;
and may posterity forget that ye were our countrymen.            - Samuel Adams

^ permalink raw reply	[flat|nested] 12+ messages in thread

* cvs cache corrupt again
  2001-11-09 10:38 cvs cache corrupt again Richard Henderson
  2001-11-09 10:47 ` Christopher Faylor
  2001-11-09 11:29 ` Angela Marie Thomas
@ 2001-12-17  9:19 ` Richard Henderson
  2 siblings, 0 replies; 12+ messages in thread
From: Richard Henderson @ 2001-12-17  9:19 UTC (permalink / raw)
  To: overseers

cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gcc/ChangeLog,v'


r~

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 10:47 ` Christopher Faylor
@ 2001-12-17  9:27   ` Christopher Faylor
  0 siblings, 0 replies; 12+ messages in thread
From: Christopher Faylor @ 2001-12-17  9:27 UTC (permalink / raw)
  To: Richard Henderson; +Cc: overseers

On Mon, Dec 17, 2001 at 09:18:22AM -0800, Richard Henderson wrote:
>cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gcc/ChangeLog,v'

I nuked fileattr.  It seems to be ok now.

cgf

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 11:29 ` Angela Marie Thomas
  2001-11-09 12:03   ` Jason Molenda
  2001-11-09 12:49   ` Ian Lance Taylor
@ 2001-12-17  9:39   ` Angela Marie Thomas
  2 siblings, 0 replies; 12+ messages in thread
From: Angela Marie Thomas @ 2001-12-17  9:39 UTC (permalink / raw)
  To: Richard Henderson; +Cc: overseers

> cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
c/ChangeLog,v'

I was told once that this happened when non-fileattr-aware versions of
cvs were used to make commits.  If that's the case, this will happen a
lot (which it seems to).  I can't think of any reason why the client would
have to know about it, so maybe it's a bug?

Chris, when you were looking at the local hacks, did you notice anything
that would indicate the client needed support for fileattr as well?

--Angela

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 12:03   ` Jason Molenda
@ 2001-12-17  9:49     ` Jason Molenda
  0 siblings, 0 replies; 12+ messages in thread
From: Jason Molenda @ 2001-12-17  9:49 UTC (permalink / raw)
  To: angela; +Cc: Richard Henderson, overseers

On Mon, Dec 17, 2001 at 09:37:40AM -0800, Angela Marie Thomas wrote:
> > cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
> c/ChangeLog,v'
> 
> I was told once that this happened when non-fileattr-aware versions of
> cvs were used to make commits.  

It's only the server that matters, and we've carefully hidden away
all unpatched cvs binaries on sourceware.  It's probably just a
bug - one of these incorrect fileattrs comes up every couple of
months.  The cvs patch could be rewritten to unlink the fileattr
cache if there is a mismatch like this.  That wouldn't be appropriate
at most sites (some other cvs server-side state can be stored in
the fileattr files), but it'd be harmless on sourceware.

Jason

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 12:49   ` Ian Lance Taylor
  2001-11-09 14:57     ` Phil Edwards
@ 2001-12-17 10:02     ` Ian Lance Taylor
  1 sibling, 0 replies; 12+ messages in thread
From: Ian Lance Taylor @ 2001-12-17 10:02 UTC (permalink / raw)
  To: angela; +Cc: overseers

Angela Marie Thomas <angela@dragaera.ReleaseDominatrix.com> writes:

> > cvs [server aborted]: head attribute does not match file for `/cvs/gcc/gcc/gc
> c/ChangeLog,v'
> 
> I was told once that this happened when non-fileattr-aware versions of
> cvs were used to make commits.  If that's the case, this will happen a
> lot (which it seems to).  I can't think of any reason why the client would
> have to know about it, so maybe it's a bug?
> 
> Chris, when you were looking at the local hacks, did you notice anything
> that would indicate the client needed support for fileattr as well?

The client doesn't need to know about the fileattrs patch.  Only the
CVS which is modifying the actual repository.

This problem can happen if a commit is interrupted at the wrong time,
or more generally if the commit fails for some reason at the wrong
time.

Ian

^ permalink raw reply	[flat|nested] 12+ messages in thread

* Re: cvs cache corrupt again
  2001-11-09 14:57     ` Phil Edwards
@ 2001-12-17 10:56       ` Phil Edwards
  0 siblings, 0 replies; 12+ messages in thread
From: Phil Edwards @ 2001-12-17 10:56 UTC (permalink / raw)
  To: Ian Lance Taylor; +Cc: angela, overseers

On Mon, Dec 17, 2001 at 10:01:50AM -0800, Ian Lance Taylor wrote:
> This problem can happen if a commit is interrupted at the wrong time,
> or more generally if the commit fails for some reason at the wrong
> time.

Yeah, somebody was talking about a modem dying mid-commit, I believe.

-- 
If ye love wealth greater than liberty, the tranquility of servitude greater
than the animating contest for freedom, go home and leave us in peace.  We seek
not your counsel, nor your arms.  Crouch down and lick the hand that feeds you;
and may posterity forget that ye were our countrymen.            - Samuel Adams

^ permalink raw reply	[flat|nested] 12+ messages in thread

end of thread, other threads:[~2001-12-17 18:56 UTC | newest]

Thread overview: 12+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2001-11-09 10:38 cvs cache corrupt again Richard Henderson
2001-11-09 10:47 ` Christopher Faylor
2001-12-17  9:27   ` Christopher Faylor
2001-11-09 11:29 ` Angela Marie Thomas
2001-11-09 12:03   ` Jason Molenda
2001-12-17  9:49     ` Jason Molenda
2001-11-09 12:49   ` Ian Lance Taylor
2001-11-09 14:57     ` Phil Edwards
2001-12-17 10:56       ` Phil Edwards
2001-12-17 10:02     ` Ian Lance Taylor
2001-12-17  9:39   ` Angela Marie Thomas
2001-12-17  9:19 ` Richard Henderson

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).