Details
Diff Detail
- Repository
- rLLD LLVM Linker
Event Timeline
Patch LGTM, please wait for Rui approval.
| ELF/Target.h | ||
|---|---|---|
| 150 | nit: I would shorten this: | |
| ELF/Target.h | ||
|---|---|---|
| 150 | Sounds good, I'll change if it to that if Rui also agrees. | |
| ELF/Target.h | ||
|---|---|---|
| 171 | Since we only call this with values <= INT64_MAX that is much better, thanks. | |
| ELF/Target.h | ||
|---|---|---|
| 171 | But why INT64_MAX? It seems just inconsistent. If you sign-extend, you want to always do that, no? | |
| ELF/Target.h | ||
|---|---|---|
| 171 | I meant that this function is either called with a negative signed number or an unsigned value that doesn't have the int64 sign bit set so that using Twine((int64_t)V) is exactly the same as the ternary expression that I had here. | |
| ELF/Target.h | ||
|---|---|---|
| 171 | If Twine((int64_t)V) is exactly the same as the ternary expression, please use the simpler form which is Twine((int64_t)V). | |
| ELF/Target.h | ||
|---|---|---|
| 171 | Yes that is exactly what I meant. Sorry about the miscommunication. | |
In general a function name should be a verb, so how about reportRangeError?