#
f3e2126c |
|
12-Sep-2017 |
Roland McGrath <mcgrathr@google.com> |
[zx] Magenta -> Zircon The Great Renaming is here! Change-Id: I3229bdeb2a3d0e40fb4db6fec8ca7d971fbffb94
|
#
151f2374 |
|
05-Sep-2017 |
Roland McGrath <mcgrathr@google.com> |
[ulib][musl] Simplify assembly alias macros A single "sym = othersym" handles .type and .size automatically. So these macros can be simpler. Change-Id: I9b7b45a8f426f7d59a68f7f5089b3664bc0fd9cd
|
#
cb5898a3 |
|
20-Mar-2017 |
Roland McGrath <mcgrathr@google.com> |
[ulib][musl] Make sigsetjmp/siglongjmp aliases for setjmp/longjmp Since we don't have any kind of signal support, there is nothing that sigsetjmp/siglongjmp need to do beyond what plain setjmp/longjmp do. The old assembly implementations had bad assumptions about the implementation details of the underlying setjmp/longjmp assembly code. MG-619 #done Change-Id: Ie1524b655f2eee24bada77bfdcdc6c83a4298bf4
|
#
01fe3baf |
|
10-Mar-2017 |
Roland McGrath <mcgrathr@google.com> |
[ulib][musl] Rewrite setjmp/longjmp The new implementation saves and restores the unsafe SP so that setjmp calls in translation units built without -fsanitize=safe-stack will be compatible with -fsanitize=safe-stack code elsewhere in their call chains. It mangles the register values that usually hold stack or code addresses stored in jmp_buf, so it's much harder to find such addresses on the heap. It includes complete CFI for longjmp so that every point unwinds correctly either to the caller of longjmp or to the caller of setjmp in the context being restored. Change-Id: I19fcb16a5e93a0219d3d840215f3695dccdb98b5
|
#
f5418c81 |
|
03-Mar-2017 |
Brian Swetland <swetland@google.com> |
[ulib][musl] rename .s files to .S This aligns with the rest of Magenta and allows for some build system cleanup and simplification. Change-Id: I2dfadccf3eaf875c3b27ed22454588e0b1fa2037
|