Breakage on master

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
7 messages Options
Reply | Threaded
Open this post in threaded view
|

Breakage on master

Ben Gamari-2
Hi all,

It seems that a patch breaking master for DEBUG-enabled builds has snuck
into `master`. I'm currently working on fixing the issue but in the
meantime the issue can be worked around by reverting
accbc242e555822a2060091af7188ce6e9b0144e.

Cheers,

- Ben

_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs

signature.asc (497 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

RE: Breakage on master

GHC - devs mailing list
|  meantime the issue can be worked around by reverting
|  accbc242e555822a2060091af7188ce6e9b0144e.

Alas, not so.

git revert accbc242e555822a2060091af7188ce6e9b0144e
warning: Failed to merge submodule utils/haddock (commits don't follow merge-base)
error: could not revert accbc242e5... DynFlags: disentangle Outputable
hint: after resolving the conflicts, mark the corrected paths
hint: with 'git add <paths>' or 'git rm <paths>'
hint: and commit the result with 'git commit'

|  -----Original Message-----
|  From: ghc-devs <[hidden email]> On Behalf Of Ben Gamari
|  Sent: 18 August 2020 17:07
|  To: GHC developers <[hidden email]>
|  Subject: Breakage on master
|  
|  Hi all,
|  
|  It seems that a patch breaking master for DEBUG-enabled builds has
|  snuck
|  into `master`. I'm currently working on fixing the issue but in the
|  meantime the issue can be worked around by reverting
|  accbc242e555822a2060091af7188ce6e9b0144e.
|  
|  Cheers,
|  
|  - Ben
_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs
Reply | Threaded
Open this post in threaded view
|

RE: Breakage on master

Ben Gamari-2
Simon Peyton Jones <[hidden email]> writes:

> |  meantime the issue can be worked around by reverting
> |  accbc242e555822a2060091af7188ce6e9b0144e.
>
> Alas, not so.
>
> git revert accbc242e555822a2060091af7188ce6e9b0144e
> warning: Failed to merge submodule utils/haddock (commits don't follow merge-base)
> error: could not revert accbc242e5... DynFlags: disentangle Outputable
> hint: after resolving the conflicts, mark the corrected paths
> hint: with 'git add <paths>' or 'git rm <paths>'
> hint: and commit the result with 'git commit'
>
Sigh, yes, this is what I was afraid of. Strangely, Hadrian's validate
flavour doesn't appear to be affected by the issue that you reported.
Are you using the make build system by any chance?

Cheers,

- Ben


_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs

signature.asc (497 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

RE: Breakage on master

Ben Gamari-2
Ben Gamari <[hidden email]> writes:

> Simon Peyton Jones <[hidden email]> writes:
>
>> |  meantime the issue can be worked around by reverting
>> |  accbc242e555822a2060091af7188ce6e9b0144e.
>>
>> Alas, not so.
>>
>> git revert accbc242e555822a2060091af7188ce6e9b0144e
>> warning: Failed to merge submodule utils/haddock (commits don't follow merge-base)
>> error: could not revert accbc242e5... DynFlags: disentangle Outputable
>> hint: after resolving the conflicts, mark the corrected paths
>> hint: with 'git add <paths>' or 'git rm <paths>'
>> hint: and commit the result with 'git commit'
>>
> Sigh, yes, this is what I was afraid of. Strangely, Hadrian's validate
> flavour doesn't appear to be affected by the issue that you reported.
> Are you using the make build system by any chance?
>
Strangely, `./validate --legacy --slow` also appears to work fine for me
on 55fd1dc55990623dcf3b2e6143e766242315d757.

Simon, can you describe how you were previously building GHC?

Cheers,

- Ben

_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs

signature.asc (497 bytes) Download Attachment
Reply | Threaded
Open this post in threaded view
|

RE: Breakage on master

GHC - devs mailing list
|  Strangely, `./validate --legacy --slow` also appears to work fine for
|  me
|  on 55fd1dc55990623dcf3b2e6143e766242315d757.
|  
|  Simon, can you describe how you were previously building GHC?

./validate --legacy --fast

with this validate.mk (below).

But the issue is terribly simple: assertPprPanic is used (by ASSERT), but no longer imported (by many many modules) because they previously got it from Outputable.  How can this possibly work?  It certainly doesn't for me.

Would it be possible to revert the patch that broke this?  I'm fully stalled with no workaround.

Thanks

Simon



SRC_HC_OPTS        = -O -H64m
GhcStage1HcOpts    = -DDEBUG
GhcStage2HcOpts    = -dcore-lint -ticky
GhcLibHcOpts       = -O -dcore-lint -ticky

BUILD_PROF_LIBS    = NO
SplitSections      = NO
HADDOCK_DOCS       = NO
BUILD_SPHINX_HTML  = NO
BUILD_SPHINX_PDF   = NO
BUILD_MAN          = NO

LAX_DEPENDENCIES   = YES


|  -----Original Message-----
|  From: Ben Gamari <[hidden email]>
|  Sent: 18 August 2020 20:04
|  To: Simon Peyton Jones <[hidden email]>; GHC developers <ghc-
|  [hidden email]>
|  Subject: RE: Breakage on master
|  
|  Ben Gamari <[hidden email]> writes:
|  
|  > Simon Peyton Jones <[hidden email]> writes:
|  >
|  >> |  meantime the issue can be worked around by reverting
|  >> |  accbc242e555822a2060091af7188ce6e9b0144e.
|  >>
|  >> Alas, not so.
|  >>
|  >> git revert accbc242e555822a2060091af7188ce6e9b0144e
|  >> warning: Failed to merge submodule utils/haddock (commits don't
|  follow merge-base)
|  >> error: could not revert accbc242e5... DynFlags: disentangle
|  Outputable
|  >> hint: after resolving the conflicts, mark the corrected paths
|  >> hint: with 'git add <paths>' or 'git rm <paths>'
|  >> hint: and commit the result with 'git commit'
|  >>
|  > Sigh, yes, this is what I was afraid of. Strangely, Hadrian's
|  validate
|  > flavour doesn't appear to be affected by the issue that you
|  reported.
|  > Are you using the make build system by any chance?
|  >
|  Strangely, `./validate --legacy --slow` also appears to work fine for
|  me
|  on 55fd1dc55990623dcf3b2e6143e766242315d757.
|  
|  Simon, can you describe how you were previously building GHC?
|  
|  Cheers,
|  
|  - Ben
_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs
Reply | Threaded
Open this post in threaded view
|

Re: Breakage on master

Sylvain Henry-2
I can't reproduce the issue. Is it on a specific branch?

If it's in a branch with new ASSERTs, you should just have to import
GHC.Utils.Panic to fix the issue or is there something else?

Sylvain

On 19/08/2020 10:24, Simon Peyton Jones via ghc-devs wrote:

> |  Strangely, `./validate --legacy --slow` also appears to work fine for
> |  me
> |  on 55fd1dc55990623dcf3b2e6143e766242315d757.
> |
> |  Simon, can you describe how you were previously building GHC?
>
> ./validate --legacy --fast
>
> with this validate.mk (below).
>
> But the issue is terribly simple: assertPprPanic is used (by ASSERT), but no longer imported (by many many modules) because they previously got it from Outputable.  How can this possibly work?  It certainly doesn't for me.
>
> Would it be possible to revert the patch that broke this?  I'm fully stalled with no workaround.
>
> Thanks
>
> Simon
>
>
>
> SRC_HC_OPTS        = -O -H64m
> GhcStage1HcOpts    = -DDEBUG
> GhcStage2HcOpts    = -dcore-lint -ticky
> GhcLibHcOpts       = -O -dcore-lint -ticky
>
> BUILD_PROF_LIBS    = NO
> SplitSections      = NO
> HADDOCK_DOCS       = NO
> BUILD_SPHINX_HTML  = NO
> BUILD_SPHINX_PDF   = NO
> BUILD_MAN          = NO
>
> LAX_DEPENDENCIES   = YES
>
>
> |  -----Original Message-----
> |  From: Ben Gamari <[hidden email]>
> |  Sent: 18 August 2020 20:04
> |  To: Simon Peyton Jones <[hidden email]>; GHC developers <ghc-
> |  [hidden email]>
> |  Subject: RE: Breakage on master
> |
> |  Ben Gamari <[hidden email]> writes:
> |
> |  > Simon Peyton Jones <[hidden email]> writes:
> |  >
> |  >> |  meantime the issue can be worked around by reverting
> |  >> |  accbc242e555822a2060091af7188ce6e9b0144e.
> |  >>
> |  >> Alas, not so.
> |  >>
> |  >> git revert accbc242e555822a2060091af7188ce6e9b0144e
> |  >> warning: Failed to merge submodule utils/haddock (commits don't
> |  follow merge-base)
> |  >> error: could not revert accbc242e5... DynFlags: disentangle
> |  Outputable
> |  >> hint: after resolving the conflicts, mark the corrected paths
> |  >> hint: with 'git add <paths>' or 'git rm <paths>'
> |  >> hint: and commit the result with 'git commit'
> |  >>
> |  > Sigh, yes, this is what I was afraid of. Strangely, Hadrian's
> |  validate
> |  > flavour doesn't appear to be affected by the issue that you
> |  reported.
> |  > Are you using the make build system by any chance?
> |  >
> |  Strangely, `./validate --legacy --slow` also appears to work fine for
> |  me
> |  on 55fd1dc55990623dcf3b2e6143e766242315d757.
> |
> |  Simon, can you describe how you were previously building GHC?
> |
> |  Cheers,
> |
> |  - Ben
> _______________________________________________
> ghc-devs mailing list
> [hidden email]
> http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs
_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs
Reply | Threaded
Open this post in threaded view
|

RE: Breakage on master

GHC - devs mailing list
|  > But the issue is terribly simple: assertPprPanic is used (by
|  ASSERT), but no longer imported (by many many modules) because they
|  previously got it from Outputable.  How can this possibly work?  

I believe that that the answer to my qn is "just import GHC.Utils.Panic".

More specifically, as part of your patch, in future, if you want to use ASSERTs you must

* BOTH #include HsVersions.h
* AND  import GHC.Utils.Panic

and that you have added that extra import to the zillion existing modules that have an ASSERT.

I had not understood that change.  I think all I need do is to add the extra import to my new modules in !3220.

It would be helpful to add a prominent comment in HsVersions.h.

I still wonder whether HsVersions.h could simply include 'import GHC.Utils.Panic' to save this clutter.  But regardless, I think I'm unstuck now.  Thanks

Simon



|  -----Original Message-----
|  From: ghc-devs <[hidden email]> On Behalf Of Sylvain
|  Henry
|  Sent: 19 August 2020 11:53
|  To: [hidden email]
|  Subject: Re: Breakage on master
|  
|  I can't reproduce the issue. Is it on a specific branch?
|  
|  If it's in a branch with new ASSERTs, you should just have to import
|  GHC.Utils.Panic to fix the issue or is there something else?
|  
|  Sylvain
|  
|  On 19/08/2020 10:24, Simon Peyton Jones via ghc-devs wrote:
|  > |  Strangely, `./validate --legacy --slow` also appears to work fine
|  for
|  > |  me
|  > |  on 55fd1dc55990623dcf3b2e6143e766242315d757.
|  > |
|  > |  Simon, can you describe how you were previously building GHC?
|  >
|  > ./validate --legacy --fast
|  >
|  > with this validate.mk (below).
|  >
|  > But the issue is terribly simple: assertPprPanic is used (by
|  ASSERT), but no longer imported (by many many modules) because they
|  previously got it from Outputable.  How can this possibly work?  It
|  certainly doesn't for me.
|  >
|  > Would it be possible to revert the patch that broke this?  I'm fully
|  stalled with no workaround.
|  >
|  > Thanks
|  >
|  > Simon
|  >
|  >
|  >
|  > SRC_HC_OPTS        = -O -H64m
|  > GhcStage1HcOpts    = -DDEBUG
|  > GhcStage2HcOpts    = -dcore-lint -ticky
|  > GhcLibHcOpts       = -O -dcore-lint -ticky
|  >
|  > BUILD_PROF_LIBS    = NO
|  > SplitSections      = NO
|  > HADDOCK_DOCS       = NO
|  > BUILD_SPHINX_HTML  = NO
|  > BUILD_SPHINX_PDF   = NO
|  > BUILD_MAN          = NO
|  >
|  > LAX_DEPENDENCIES   = YES
|  >
|  >
|  > |  -----Original Message-----
|  > |  From: Ben Gamari <[hidden email]>
|  > |  Sent: 18 August 2020 20:04
|  > |  To: Simon Peyton Jones <[hidden email]>; GHC developers
|  <ghc-
|  > |  [hidden email]>
|  > |  Subject: RE: Breakage on master
|  > |
|  > |  Ben Gamari <[hidden email]> writes:
|  > |
|  > |  > Simon Peyton Jones <[hidden email]> writes:
|  > |  >
|  > |  >> |  meantime the issue can be worked around by reverting
|  > |  >> |  accbc242e555822a2060091af7188ce6e9b0144e.
|  > |  >>
|  > |  >> Alas, not so.
|  > |  >>
|  > |  >> git revert accbc242e555822a2060091af7188ce6e9b0144e
|  > |  >> warning: Failed to merge submodule utils/haddock (commits
|  don't
|  > |  follow merge-base)
|  > |  >> error: could not revert accbc242e5... DynFlags: disentangle
|  > |  Outputable
|  > |  >> hint: after resolving the conflicts, mark the corrected paths
|  > |  >> hint: with 'git add <paths>' or 'git rm <paths>'
|  > |  >> hint: and commit the result with 'git commit'
|  > |  >>
|  > |  > Sigh, yes, this is what I was afraid of. Strangely, Hadrian's
|  > |  validate
|  > |  > flavour doesn't appear to be affected by the issue that you
|  > |  reported.
|  > |  > Are you using the make build system by any chance?
|  > |  >
|  > |  Strangely, `./validate --legacy --slow` also appears to work fine
|  for
|  > |  me
|  > |  on 55fd1dc55990623dcf3b2e6143e766242315d757.
|  > |
|  > |  Simon, can you describe how you were previously building GHC?
|  > |
|  > |  Cheers,
|  > |
|  > |  - Ben
|  > _______________________________________________
|  > ghc-devs mailing list
|  > [hidden email]
|  >
|  https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fmail.
|  haskell.org%2Fcgi-bin%2Fmailman%2Flistinfo%2Fghc-
|  devs&amp;data=02%7C01%7Csimonpj%40microsoft.com%7Cee9f099b376042e0fb6c
|  08d8442e2b39%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637334312373
|  953419&amp;sdata=6XaNrQIfRNqlSnXPAmoxXYTYQorYszHmyJc1kThIU2o%3D&amp;re
|  served=0
|  _______________________________________________
|  ghc-devs mailing list
|  [hidden email]
|  https://nam06.safelinks.protection.outlook.com/?url=http%3A%2F%2Fmail.
|  haskell.org%2Fcgi-bin%2Fmailman%2Flistinfo%2Fghc-
|  devs&amp;data=02%7C01%7Csimonpj%40microsoft.com%7Cee9f099b376042e0fb6c
|  08d8442e2b39%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637334312373
|  963414&amp;sdata=m2rdARx2tpZFv5I2WiEdYE5d37LeXTMIIPg51Zfee9M%3D&amp;re
|  served=0
_______________________________________________
ghc-devs mailing list
[hidden email]
http://mail.haskell.org/cgi-bin/mailman/listinfo/ghc-devs