The instruction can be 16-bit aligned while targeting 32-bit aligned code. To calculate the target address correctly, the address of the instruction has to be adjusted.
Details
Details
Diff Detail
Diff Detail
- Repository
- rG LLVM Github Monorepo
Event Timeline
Comment Actions
Change looks good to me. There is some usage of BLX in the LLD test suite. If you've not done so already it will be worth doing a check-all including the LLD test suite to make sure that there are no tests that need updating.
Comment Actions
Thanks for looking at it.
As far as I can see, such tests either have the BLX instructions 32-bit aligned or check only their immediates.