- User Since
- Oct 28 2014, 11:11 AM (364 w, 6 d)
Mar 10 2021
Mar 8 2021
Aug 7 2016
The file was auto-formatted via clang-format on save, via Atom. Sorry about the noise.
Nov 12 2015
Nov 10 2015
Nov 4 2015
It's actually an extra newline character that my editor automatically stripped out. *nix has the requirement that all lines must end with a newline character, not that a file must end with an empty line.
Okay, so run this on the current HEAD:
$ cmake -GNinja -DCMAKE_BUILD_TYPE=Debug ../lldb -- The C compiler identification is AppleClang 22.214.171.12420053 -- The CXX compiler identification is AppleClang 126.96.36.19920053 -- Check for working C compiler using: Ninja -- Check for working C compiler using: Ninja -- works -- Detecting C compiler ABI info -- Detecting C compiler ABI info - done -- Detecting C compile features -- Detecting C compile features - done -- Check for working CXX compiler using: Ninja -- Check for working CXX compiler using: Ninja -- works -- Detecting CXX compiler ABI info -- Detecting CXX compiler ABI info - done -- Detecting CXX compile features -- Detecting CXX compile features - done CMake Error at cmake/modules/LLDBStandalone.cmake:37 (get_filename_component): get_filename_component called with incorrect number of arguments Call Stack (most recent call first): CMakeLists.txt:3 (include)
Nov 3 2015
Pretty silly mistake on my part. Thanks for the clear tests.
Oct 16 2015
Er, silly mistake (build failure)
Aug 27 2015
LGTM, with some minor doubts.
Feb 25 2015
Feb 14 2015
I submitted the relevant test with 69a5c891.
Feb 9 2015
Don't assume that the derived pointer can always be cast to an
Argument; dyn_cast and check.
Do try to rebase onto my changes; I did spend quite a bit of time painfully splitting the patches.
As requested by reames, shrink the original submission to be just the
bugfix. Other changes will come as later patches.
Feb 8 2015
Drop the deref propagation, because it requires more work. Redo the
nonNull propagation so that it is correct (thanks to Hal for poking me
to test it properly).
A quick run through clang-format.
More work in the same direction.
Feb 5 2015
ping! (How) should I write a test for this?
Jan 27 2015
Split print-memderefs into its own change. Fix all-round sloppiness.
Jan 23 2015
Abandoning: not important enough to wait for; this test file can be rewritten to exercise Any comprehensively, if anyone's interested.
Abandoning: it's not important enough; I have no dependent changes.
Jan 22 2015
This should be really easy to review and get out of the way; it's borderline non-functional.
Jan 21 2015
I have too many changes awaiting review to continue working :\
Jan 20 2015
Jan 18 2015
Elena Demikhovsky wrote:
When you put "any", you probably mean any scalar of int, float, ptr. What happens if somebody specifies vector, i16, double or something else. May be you should put some code in the verifier?
Jan 17 2015
Jan 16 2015
Don't remove old intrinsics right away; just deprecate them.
Jan 15 2015
Rebase to include statepoint-example gc.
After reading through StatepointLowering in greater detail, I realize that I should be testing the stackmaps, not the generic stores that are emitted; emitting those stores is not part of the statepoint code.
Philip Reames wrote:
As I said before, LGTM.