| Mercury Bugs - mercury | |||||
| View Issue Details | |||||
| ID | Project | Category | View Status | Date Submitted | Last Update | 
| 0000128 | mercury | Bug | public | 2010-01-14 18:19 | 2010-09-15 12:29 | 
| Reporter | juliensf | ||||
|---|---|---|---|---|---|
| Assigned To | zs | ||||
| Priority | normal | Severity | minor | Reproducibility | have not tried | 
| Status | resolved | Resolution | fixed | ||
| Platform | OS | OS Version | |||
| Product Version | |||||
| Target Version | Fixed in Version | ||||
| Summary | 0000128: merge_instmapping_delta_2: error merging var 20 | ||||
| Description | (This was originally reported by Michael Day on the mercury-bugs list; responses from zs are included below as well.) Hi, The attached program triggers a compiler error when building with -O5: Making Mercury/int3s/bug.int3 Making Mercury/ints/bug.int Making Mercury/cs/bug.c Uncaught Mercury exception: Software Error: instmap.m: Unexpected: merge_instmapping_delta_2: error merging var 20 It succeeds at lower optimisation levels. Tested with rotd-2009-10-08 and rotd-2008-01-30. Michael ============ Date: Mon, 9 Nov 2009 17:43:53 +1100 From: Zoltan Somogyi <zs@csse.unimelb.edu.au> To: Michael Day <mikeday@yeslogic.com> Cc: mercury-bugs@csse.unimelb.edu.au Subject: Re: merge_instmapping_delta_2: error merging var 20 On 09-Nov-2009, Michael Day <mikeday@yeslogic.com> wrote: > The attached program triggers a compiler error when building with -O5: > > Making Mercury/int3s/bug.int3 > Making Mercury/ints/bug.int > Making Mercury/cs/bug.c > Uncaught Mercury exception: > Software Error: instmap.m: Unexpected: merge_instmapping_delta_2: error > merging var 20 > > It succeeds at lower optimisation levels. Tested with rotd-2009-10-08 and > rotd-2008-01-30. It also succeeds with -O5 --no-loop-invariants, and the exception is thrown *during* the loop invariants pass. I think that makes this Ralph's to fix. Zoltan. ======= Hi Zoltan, > It also succeeds with -O5 --no-loop-invariants, and the exception is thrown > *during* the loop invariants pass. I think that makes this Ralph's to fix. Suits me! :) The bug can be worked around by creating lots of little helper predicates, although it is a bit tedious. We always build with -O5 rather than waiting until just before release, as it's very frustrating to suddenly be hit with an unexpected internal compiler error. Best regards, Michael -- Print XML with Prince! http://www.princexml.com | ||||
| Tags | No tags attached. | ||||
| Relationships | |||||
| Attached Files |  bug.m (1,078) 2010-01-14 18:20 https://bugs.mercurylang.org/file_download.php?file_id=93&type=bug | ||||
| Notes | |||||
|  | |||||
| 
 | 
 | ||||
| Issue History | |||||
| Date Modified | Username | Field | Change | ||
|---|---|---|---|---|---|
| 2010-01-14 18:19 | juliensf | New Issue | |||
| 2010-01-14 18:20 | juliensf | File Added: bug.m | |||
| 2010-09-14 19:47 | zs | Status | new => assigned | ||
| 2010-09-14 19:47 | zs | Assigned To | => zs | ||
| 2010-09-15 12:29 | zs | Note Added: 0000290 | |||
| 2010-09-15 12:29 | zs | Status | assigned => resolved | ||
| 2010-09-15 12:29 | zs | Resolution | open => fixed | ||