Bug 20220 - [4.0/4.1 Regression] Rejecting invalid template code, breaks most of Boost
Summary: [4.0/4.1 Regression] Rejecting invalid template code, breaks most of Boost
Status: RESOLVED FIXED
Alias: None
Product: gcc
Classification: Unclassified
Component: c++ (show other bugs)
Version: 4.0.0
: P2 critical
Target Milestone: 4.0.0
Assignee: Not yet assigned to anyone
URL:
Keywords: rejects-valid
: 20261 (view as bug list)
Depends on:
Blocks:
 
Reported: 2005-02-26 00:44 UTC by Giovanni Bajo
Modified: 2005-03-01 14:32 UTC (History)
4 users (show)

See Also:
Host:
Target:
Build:
Known to work:
Known to fail:
Last reconfirmed: 2005-02-26 00:55:09

Attachments
Preprocessed source (to be reduced) (86.18 KB, application/octet-stream)
2005-02-26 00:46 UTC, Giovanni Bajo 		Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Giovanni Bajo 2005-02-26 00:44:33 UTC 
The attached preprocessed source code is an example of a recent regresion of 
the C++ frontend which totally *kills* Boost: test rate dropped from 90% to 9% 
because of this! This used to work like 2 weeks ago...
Comment 1 Giovanni Bajo 2005-02-26 00:46:01 UTC 
Created attachment 8288 [details]
Preprocessed source (to be reduced)
Comment 2 Andrew Pinski 2005-02-26 00:55:09 UTC 
Easy reduction:
template <class T, T i> class b {};
template <class T, T i> class a
{
static const T value = i;
  b<T, static_cast<T>(value+1) > next;
};
Comment 3 Andrew Pinski 2005-02-26 00:58:02 UTC 
This worked with 4.0.0 20050113 but fails with 20050210.
Comment 4 Giovanni Bajo 2005-02-26 12:01:50 UTC 
I forgot to thank Martin Wille for notifying me of this bug and providing me 
with this testcase.
Comment 5 Andrew Pinski 2005-02-28 01:09:39 UTC 
(In reply to comment #3)
> This worked with 4.0.0 20050113 but fails with 20050210.
To get a smaller window, it fails with 20050201.
Hmm, it passes with 20050225.
I think this was fixed by:
        PR c++/19991
        * init.c (integral_constant_value): Iterate if the value of a decl
        is itself a constant.

Since this is fixed both the orginal full test case and the reduced version, I am going to close it as fixed.
Giovanni if you want to apply the testcase I think that would be useful for the future.
Comment 6 Andrew Pinski 2005-03-01 14:32:39 UTC 
*** Bug 20261 has been marked as a duplicate of this bug. ***