Messages in this thread | | | Date | Fri, 6 Apr 2018 19:03:16 +0200 | From | Ingo Molnar <> | Subject | Re: [PATCH 0/8] use struct pt_regs based syscall calling for x86-64 |
| |
* Dominik Brodowski <linux@dominikbrodowski.net> wrote:
> > I.e. I'd generate the names like this: > > > > __{x64,x32,ia32}[_compat]_sys_waittid() > > > > The fully consistent nomenclature would be someting like this: > > > > ffffffff8105f1e0 t kernel_waitid # common C function (64-bit kargs) > > ffffffff8105f2b0 t SYS_waitid # 64-bit uaddr args C function > > ffffffff8105f410 T __x64_sys_waitid # 64-bit-ptregs -> C stub > > ffffffff8105f430 T __ia32_sys_waitid # 32-bit-ptregs -> C stub > > ffffffff8105f450 t COMPAT_SYS_waitid # 32-bit uaddr args C function > > ffffffff8105f5e0 T __ia32_compat_sys_waitid # 32-bit-ptregs -> C stub > > ffffffff8105f600 T __x32_compat_sys_waitid # 64-bit-ptregs -> C stub > > > > Looks a lot tidier and a lot more logical, doesn't it? > > Indeed. Want me to prepare a new patch 8/8 on top which does the renaming > (for x86 and for the generic case), or will you do the re-naming while > merging my patches yourself?
Please do an 8/8 patch that does the rename - I'll push out the first 7 patches so they get more testing.
Note, I have not checked the above name space for namespace collisions - but unless we are unlucky it should be fine.
BTW., is there any deep reason why some of these names are capitalized?
I.e. could we use:
ffffffff8105f1e0 t kernel_waitid # common C function (64-bit kargs) ffffffff8105f2b0 t sys_waitid # 64-bit uaddr args C function ffffffff8105f410 T __x64_sys_waitid # 64-bit-ptregs -> C stub ffffffff8105f430 T __ia32_sys_waitid # 32-bit-ptregs -> C stub ffffffff8105f450 t compat_sys_waitid # 32-bit uaddr args C function ffffffff8105f5e0 T __ia32_compat_sys_waitid # 32-bit-ptregs -> C stub ffffffff8105f600 T __x32_compat_sys_waitid # 64-bit-ptregs -> C stub
?
Note how this reduces naming complexity and increases the self-consistency even more.
Thanks,
Ingo
| |