- User Since
- Oct 29 2016, 4:19 PM (221 w, 4 d)
Nov 15 2020
Jul 1 2020
I'd like some input on whether or not to include fixup dumping in the -all option. Relocations can quickly become numerous in a moderately complex binary.
Jun 23 2020
Thank you! I do not have commit access, so I will need some help with that.
Jun 22 2020
- Added a test for scoped enums (works as-is but still worth testing).
- Less frivolous use of auto
- Made the floating point vars constexpr to maintain validity in clang's internals.
- AstRestoreTest CLASS tests run for both DIA and Native. A pure NativePDB test that does not depend on MSVC would require extending s_constant.cpp (S_CONSTANT not generated by clang-cl) but I'm not certain how to regenerate the listing file with comments (unless that's done by hand). For now, this change should only be considered relevant for PDBs generated by MSVC.
Jun 21 2020
Apply formatting fixes
Fix ternary to use corresponding float/double overloaded constructors of APFloat
Jun 20 2020
Review changes from @teemperor.
Jun 19 2020
Jan 15 2018
I've found that passing --gcc-toolchain=/usr results in the exact behavior I'm after.
Dec 22 2017
Dec 21 2017
Jan 27 2017
Remove already-aliased option matchings; add test case for patch.
Nov 30 2016
Rebase against upstream master. Cleanup PPCTargetInfo::relocateOne() with mask constants.
Rebase against upstream master
Rebase against upstream master
Nov 21 2016
Add target-independent small data classification for constants via MachineConstantPoolEntry::getSectionKind().
Nov 18 2016
Implement getBitcodeOSABIKind, which just responds to Triple::EABI for now.
Nov 17 2016
Test now uses sda21 symbol variant as changed from sdarx.
To better follow GCC convention, use sda21 rather than sdarx as the symbol variant.
Ensure Config->OSABI == ELF::ELFOSABI_STANDALONE is true when synthesizing _SDA_BASE_ symbols.
Object file ABI tag is now set with the following for PPC:
Nov 16 2016
Make PPC target bi-endian in the same manner as MIPS. Tidy up PPCTargetInfo::relocateOne(). Rebase with latest master
Rebase against latest master.
Rebase against latest master. check-all test results match those experienced on base master.
Nov 12 2016
Add PPCTargetInfo::relocateOne() R_PPC_ADDR14 and R_PPC_REL14 handling.
_SDA_BASE_ symbols need direct calls to ElfSym<ELFT>::addSynthetic to be initially created.
Nov 11 2016
Rebase with upstream master. Add -sda-base= and -sda2-base= linker options for overriding synthetic values.
Nov 10 2016
Remove stray blank line, Rebase against upstream master.
Remove PPCSubtarget:: useEABISmallDataSections() flag, opting for direct TargetTriple::isEABI().
Nov 9 2016
Little-endian addressing fix. Uses addressing described in "188.8.131.52 PowerPC Instruction Addressing in Little-Endian Mode" section of programming manual. (needed to think of it from the offset operand end)
Create SmallKind = 0x80 enum entry, rename smallUnderlyingKind() to UnderlyingKind()
Nov 6 2016
Committed by ruiu as part of r285720.
Nov 1 2016
All I've set up is the Phabricator access, no commit privileges to my knowledge
Oct 31 2016
Commit R_PPC_ADDR32 implementation in tandem with test
Ensure patch contents reflects title (no ADDR32)
Restore LGTM'd patch
Just to clarify, update diff to include *just* REL24/REL32, then submit ADDR32 standalone?
Since ADDR32 was a trivial relocation for the last diff, this test is added to validate
Oct 30 2016
This adds a test for REL24.
Oct 29 2016
That ends up emitting R_PPC_ADDR32 (absolute relocation, which also needs to be implemented).
Ok, I've appended a simple test for REL24 in ppc-relocs.s: