On Tue, 2013-12-10 at 15:12 -0800, Jordan Justen wrote: > > I don't think Andrew was suggesting that we commit a change that would > break CLANG builds. Particularly when building BaseLib which is used > by basically every EDK II image.
This would not break CLANG builds. I already reported that the Thunk16
code would be miscompiled by CLANG, but of course that left the
possibility that rest of MdePkg might have at least *built*, and people
without a CSM would never have cared.
That isn't the case, though. The existing code doesn't even build
*incorrectly* with CLANG, since LLVM doesn't admit the existence of the
16-bit forms of the pusha/popa instructions, and thus dies with the
following errors:
Thunk16.iii:43:5: error: invalid instruction mnemonic 'pushaw'
pushaw # pushad actually
~~~~^~~~~~~~~~~~~~~~~~~~~~~~
Thunk16.iii:112:5: error: invalid instruction mnemonic 'popaw'
popaw # popad actually
^
Ironic, really, that we actually want the 32-bit forms of the
instructions here anyway :)
This code was *already* uncompilable with CLANG, and would result in
build errors. Having it die with ".code16 not supported yet" would be an
*improvement*.
--
David Woodhouse Open Source Technology Centre
david.woodho...@intel.com Intel Corporation
smime.p7s
Description: S/MIME cryptographic signature
------------------------------------------------------------------------------ Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk
_______________________________________________ edk2-devel mailing list edk2-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/edk2-devel
