2011/10/24 Simon Peyton-Jones :
> | The bad thing is that the .haddock file exists, but is from an earlier
> | version of haddock. haddock doesn't like that at all. Or rather, it doesn't
> | notice and misinterprets it, leading to a segfault.
>
> I think that's exactly right. Sometimes you get "Ix
> | The bad thing is that the .haddock file exists, but is from an earlier
> | version of haddock. haddock doesn't like that at all. Or rather, it
> | doesn't notice and misinterprets it, leading to a segfault.
>
> I think that's exactly right. Sometimes you get "Ix out of range", and
> sometimes
| The bad thing is that the .haddock file exists, but is from an earlier
| version of haddock. haddock doesn't like that at all. Or rather, it doesn't
| notice and misinterprets it, leading to a segfault.
I think that's exactly right. Sometimes you get "Ix out of range", and
sometimes a seg fault
On Friday 21 October 2011, 23:58:59, David Waern wrote:
>
> Thanks for this information. I don't have access to a 32 bit system,
> but I've noticed that there has been changes to the .haddock file
> format in recent commits, but no update of the .haddock file version
> number. So deserialization o
2011/10/21 Daniel Fischer :
> On Friday 21 October 2011, 21:45:23, you wrote:
>>
>> Thanks Daniel and everyone else for investigating. Can you tell me
>> which version of Haddock you're running and which version the the
>> .haddock file comes from?
>
> It's surfaced during validate builds, so the v
On Friday 21 October 2011, 17:05:56, Ian Lynagh wrote:
> On Fri, Oct 21, 2011 at 01:57:07PM +, Simon Peyton-Jones wrote:
> > All I'm saying is that haddock seg-faults when given the command line
> > I gave in my message. Surely that's a bug?
>
> Yup - probably in GHC rather than haddock. I st
On Thursday 20 October 2011, 19:17:02, Ian Lynagh wrote:
> On Thu, Oct 20, 2011 at 02:25:40PM +, Simon Peyton-Jones wrote:
> > Well I can confidently say that it's the invocation of haddock that
> > seg-faults.
>
> Oh, sure, but I don't think the grep failure is related.
Both, it and the offe
On Friday 21 October 2011, 15:57:07, Simon Peyton-Jones wrote:
> All I'm saying is that haddock seg-faults when given the command line I
> gave in my message. Surely that's a bug?
Definitely. But it's hard to pin when normally you get an orderly
$ ../inplace/bin/haddock --gen-index --gen-content
Simon M, are you able to reproduce it?
Thanks
Ian
> | -Original Message-
> | From: cvs-ghc-boun...@haskell.org [mailto:cvs-ghc-boun...@haskell.org] On
> Behalf Of
> | Ian Lynagh
> | Sent: 20 October 2011 18:17
> | To: cvs-ghc@haskell.org
> | Subject: Re: Seg fault i
o: cvs-ghc@haskell.org
| Subject: Re: Seg fault in haddock
|
| On Thu, Oct 20, 2011 at 02:25:40PM +, Simon Peyton-Jones wrote:
| > Well I can confidently say that it's the invocation of haddock that
seg-faults.
|
| Oh, sure, but I don't think the grep failure is related.
|
| Howe
On Thu, Oct 20, 2011 at 02:25:40PM +, Simon Peyton-Jones wrote:
> Well I can confidently say that it's the invocation of haddock that
> seg-faults.
Oh, sure, but I don't think the grep failure is related.
However, validate has built fine for me on Windows/x86, and twice on
Linux/x86. I've al
m: cvs-ghc-boun...@haskell.org [mailto:cvs-ghc-boun...@haskell.org] On
| Behalf Of Ian Lynagh
| Sent: 20 October 2011 14:11
| To: cvs-ghc@haskell.org
| Subject: Re: Seg fault in haddock
|
|
| Hi Simon,
|
| On Thu, Oct 20, 2011 at 10:23:31AM +, Simon Peyton-Jones wrote:
| >
| >
On Thursday 20 October 2011, 15:11:29, Ian Lynagh wrote:
> Hi Simon,
>
> On Thu, Oct 20, 2011 at 10:23:31AM +, Simon Peyton-Jones wrote:
> > Note the bogus path.It is this command invocation that seg-faults.
> > (I wish the gen_contents script echoed its commands before executing
> > them
Hi Simon,
On Thu, Oct 20, 2011 at 10:23:31AM +, Simon Peyton-Jones wrote:
>
> Note the bogus path.It is this command invocation that seg-faults. (I
> wish the gen_contents script echoed its commands before executing them; could
> it be made to do so?)
>
>
> *Now, it seems th
Ian, Simon, David
Here's something that happened to me in my Windows build of GHC:
cd libraries && sh gen_contents_index --inplace
grep: Cabal/Cabal.cabal: No such file or directory
Segmentation fault/access violation in generated code
make[1]: *** [libraries/index.html] Error 1
make: *** [a
15 matches
Mail list logo
