by using the int-to-enum casting instead. The port should now build on
5.3 and wherever else the self-tests were failing before.
Submitted to vendor.
Obtained from: Michael Cornelius
ICU):
PR: 102629
Make sure to use not only "-O2", but also "-fno-strict-aliasing"
-- but only when using the "stock" cc/c++ as the newer gcc-4.x does not
seem to exhibit a problem. This seems to help the test-exposed
miscompilations.
PR: 103255
threads is broken on ia64 (marcel is investigating) and seems useless
elsewhere.
Filter `-funroll-loops' out of CFLAGS for now -- it seems to cause
miscompilation occasionally.
Correct the dependencies to stop requiring a particular version number
of the ICU's libraries to ease future upgrades.
than it does not and a thread-guru should take a closer look.
Modify the port to make `intltest' part optional, while running through
other vendor's tests. Vendor is notified --
http://www.jtcsv.com/cgibin/icu-bugs/incoming?id=1141
While here, change echo to ${ECHO}.
Add Ukrainian KOI8-U encoding and fix Ukrainian spell-out rules.
Build the vendor's tests and run them automatically after build.
Once the wrinkles (if any) are ironed out, the current users of
icu2 should switch to icu and icu2 deleted.
PR: ports/71790
Submitted by: Andriy Gapon
architectures that they need to upgrade. (Since the 2.x version has
been out for a couple of years, this port should probably just be
removed unless anyone sees a reason not to.)
