[Issue 16679] prefetch on old pentium d results in an illegal instruction

2017-01-16 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

--- Comment #7 from github-bugzi...@puremagic.com ---
Commits pushed to newCTFE at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/a5bffa8b35bb519ca7a5cd41fee86f541a1c2e76
fix Issue 16679 - prefetch on old pentium d results in an illegal instruction

https://github.com/dlang/dmd/commit/f5a73f36360fc5b2b2e93f2dd17082c4fed2b0bc
Merge pull request #6266 from WalterBright/fix16679

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-12-27 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

--- Comment #6 from github-bugzi...@puremagic.com ---
Commits pushed to scope at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/a5bffa8b35bb519ca7a5cd41fee86f541a1c2e76
fix Issue 16679 - prefetch on old pentium d results in an illegal instruction

https://github.com/dlang/dmd/commit/f5a73f36360fc5b2b2e93f2dd17082c4fed2b0bc
Merge pull request #6266 from WalterBright/fix16679

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-21 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

Walter Bright  changed:

   What|Removed |Added

 Status|REOPENED|RESOLVED
 Resolution|--- |FIXED

--- Comment #5 from Walter Bright  ---
(In reply to Brad Roberts from comment #4)
> Reopened and lowered to major from regression until the cpuid work is done
> and the test is re-enabled.  I would have suggested waiting for the right
> fix rather than the quick fix, but in neither case should we loose track of
> the disabled test.

This is already covered by https://issues.dlang.org/show_bug.cgi?id=16704

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-20 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

Brad Roberts  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Resolution|FIXED   |---
   Severity|regression  |major

--- Comment #4 from Brad Roberts  ---
Reopened and lowered to major from regression until the cpuid work is done and
the test is re-enabled.  I would have suggested waiting for the right fix
rather than the quick fix, but in neither case should we loose track of the
disabled test.

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-20 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

github-bugzi...@puremagic.com changed:

   What|Removed |Added

 Status|NEW |RESOLVED
 Resolution|--- |FIXED

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-20 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

--- Comment #3 from github-bugzi...@puremagic.com ---
Commits pushed to master at https://github.com/dlang/dmd

https://github.com/dlang/dmd/commit/a5bffa8b35bb519ca7a5cd41fee86f541a1c2e76
fix Issue 16679 - prefetch on old pentium d results in an illegal instruction

https://github.com/dlang/dmd/commit/f5a73f36360fc5b2b2e93f2dd17082c4fed2b0bc
Merge pull request #6266 from WalterBright/fix16679

fix Issue 16679 - prefetch on old pentium d results in an illegal ins…

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-20 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

Walter Bright  changed:

   What|Removed |Added

   See Also||https://issues.dlang.org/sh
   ||ow_bug.cgi?id=16704

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-20 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

Walter Bright  changed:

   What|Removed |Added

   Keywords||SIMD

--- Comment #2 from Walter Bright  ---
https://github.com/dlang/dmd/pull/6266

--

[Issue 16679] prefetch on old pentium d results in an illegal instruction

2016-11-10 Thread via Digitalmars-d-bugs 
https://issues.dlang.org/show_bug.cgi?id=16679

Brad Roberts  changed:

   What|Removed |Added

   Assignee|nob...@puremagic.com|bugzi...@digitalmars.com

--- Comment #1 from Brad Roberts  ---
The exact commits across the three repositories I've used, the druntime and
phobos commits could likely be a range of valid enough points, but these are
what I used.

dmd passes: 9a935f68feb360f78ef32913053ce05c838fcafe
dmd fails:  0fc8be57bb0107a718ac2dabd1b5482c1ebe89ef

druntime:   565b5c3163ae93858702daab9bccfcd7a4d3f748
phobos: 30aca423cdd6d7943bb1dde45f526a6078a024ae

Chances are that this is largely due to the age of the cpu given that all the
other systems running test builds are passing.  This machine is old-farm-1,
part of the fleet of auto test machines, but I've had it offline for a while
for unrelated reasons.  Unfortunate timing.  I can't add it back with the test
failing.

--