On Fri, Nov 11, 2022 at 10:31:00AM +0000, Luis Machado wrote: > On 11/1/22 14:33, Mark Brown wrote: > > Implement support for a new note type NT_ARM64_ZT providing access to > Should we use NT_AARCH64_ZT instead? This would align with our previous choice of PT_AARCH64_MEMTAG_MTE. > And we have a precedent with PT_AARCH64_ARCHEXT as well. > > #define NT_ARM_ZA 0x40c /* ARM SME ZA registers */ > > +#define NT_ARM_ZT 0x40d /* ARM SME ZT registers */ > I guess the above comment means this would also need to be NT_AARCH64_ZT, but historically we've been using NT_ARM_* even for AARCH64. > So I suppose this is OK. Indeed - there's also a lot of NT_ARM_ defines, including all the other FP defines (NT_ARM_SVE, NT_ARM_SSVE and NT_ARM_ZA) and PAC. It feels better to be consistent with at least the other FP stuff. It's only the MTE define and the .note.gnu.property stuff that's used AARCH64, and things like SVE are fairly old.
Attachment:
signature.asc
Description: PGP signature
