On Mon, Aug 8, 2022 at 12:39 AM Randy Dunlap <rdunlap@xxxxxxxxxxxxx> wrote:
> On 8/7/22 12:44, Arnd Bergmann wrote:
> > On Sun, Aug 7, 2022 at 7:28 PM Randy Dunlap <rdunlap@xxxxxxxxxxxxx> wrote:
> >>
> >
> > This does not work: __ARCH_WANT_COMPAT_FADVISE64_64 is defined in
> > arch/riscv/include/asm/unistd.h, which is not a UAPI header. By making the line
> > conditional on this, user space no longer sees the macro definition.
> >
> > It looks like you also drop the native definition on all architectures other
> > than riscv here. What we probably want is to just make all the
> > declarations in include/linux/compat.h unconditional and not have them
> > depend on architecture specific macros. Some of these may have
> > incompatible prototypes depending on the architecture, but if we run
> > into those, I would suggest we just give them unique names.
>
> Thanks for the comments.
>
> With the other patch to kernel/sys_ni.c, this one is no longer needed,
Ok.
> although I can look into making more entries in <linux/compat.h>
> unconditional.
This would be a nice cleanup, but it does involve making sure that
all prototypes are compatible with the implementation on each
architecture. I think we should definitely do this, it's just not as
simple as removing each #ifdef in linux/compat.h and linux/syscall.h
> That would also mean adding them to kernel/sys_ni.c, right?
> (if not already there)
That part should be completely independent. If the entry in
kernel/sys_ni.c is missing, that causes a link failure, while an
incorrect #ifdef would cause a compile-time error for the missing
prototype.
Arnd
