Adds iOS 6.0, which is now the default.
- changed the default iOS SDK to version 6.0
- use xcrun to locate the selected compiler
- added armv7s as an option for architecture
- made armv7 + armv7s the default architecture
- renamed the LLVM compiler option to LLVM_GNU for clarity
since LLVM is naturally associated with CLANG
This undoes commit 11565, "workaround for archaic T202 g++ bug"
which doesn't actually seem to be necessary for T202 compilation.
It also replaces 'typename' with 'class', for another compiler.
Modification provided by Dirk Zimoch.
Undo strncmp+strncpy to strcmp+strcpy change in r12048.
The 'n' versions are needed because device support was
never explicitly required to Nil-terminate the string.
The dbAccess routines do enforce termination though.
* Changing the number of elements in a VALx array triggers monitors
* Post monitors on the NEVx fields, following EFLG
* Remove artificial limit on array sizes
camonitor was using strlen() to find out the length of the array to print as %s string,
which led to printing old buffer contents when the array-string was not null terminated.
Now uses the minimum of strlen(), elements received, and elements requested.
Suggested by Mark Rivers on tech-talk (11 Sep 2012)
An internal buffer was allocated one char too short, when caput was used
with the '-S' (send string as array of chars) option.
Reported by J. Lewis Muir (tech-talk on 17-Aug-2012)
Don't add extraneous characters when SHRLIB_VERSION or GNU_TARGET
are empty. The addprefix or addsuffix functions only include the
prefix/suffix part if the list argument is not empty.
This fixes issues naming libCap5.so at build-time.
All are now robust against overflow, NAN or negative argument.
Passing seconds=0 calls the OS scheduler, offering to yield.
Passing seconds>0 delays at least the requested time, up to a
limit which usually depends on the OS tick rate.
Now works properly on Darwin where the version number comes
before the .dylib extension. This also makes Windows-specific
conditional tests fully generic.
This is really a workaround for a Win32/MinGW bug in sscanf, which
will only set the %n argument for the format "%u %n" when there is
a space following the unsigned number.
This changes the IOC's behavior very slightly.
Andrew Johnson