[Bug 1001545] ustl: update to uSTL 1.6

[bugzilla-daemon@bugs.ecos.sourceware.org]

Please do not reply to this email. Use the web interface provided at:
http://bugs.ecos.sourceware.org/show_bug.cgi?id=1001545

--- Comment #2 from Sergei Gavrikov <sergei.gavrikov@gmail.com> 2012-03-25 17:17:43 BST ---
(In reply in comment #1)
> Sergei, great to see this patch. Thank you.
> 
> The patch looks clean. You have replaced the NextPow2 implementation
> with the latest from upstream which is good. Our efforts to minimise
> the differences between eCos uSTL sources and upstream are working
> well for us.
> 
> If the tests are building and running OK with GCC 4.3.2 tools (default
> template + ustl + fileio), then I see no reason why you should not
> check this in. It is always good to check uSTL with a target
> architecture other than x86 since Mike Sharov works mainly with x86.
> We know that a few of the tests are interactive and will therefore not
> run.

Yep. I used default+fileio+ustl without any tweaks and run the tests
under GDB. A summary

- No new warnings with GCC 4.3.2 & GCC 4.6.3
- All bvt tests were run on ARMv4 (LE) target (GCC 4.3.2 & 4.6.3 builds)
  without fails
- All bvt tests (but bvt23) also were run on eCos Linux synthetic target

NOTE: it was not possible to build `bvt23' for synthetic target, but
      that is old "lrint*" issue:

      http://sourceware.org/ml/ecos-patches/2010-06/msg00027.html

-- 
Configure bugmail: http://bugs.ecos.sourceware.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.