On Tue, Nov 17, 2026 at 03:04:49PM +0100, Christophe Leroy (CS GROUP) wrote:
> Le 17/11/2026 à 13:33, Ritesh Harjani a écrit :
> > Mike Rapoport <[email protected]> writes:
> > 
> > > > > > Hence, use the modules area for EXECMEM_KPROBES on book3s/32.
> > > > > Right, I missed that on book3s/32 does vmalloc is not executable and 
> > > > > its
> > > > > protection can't be changed.
> > > > > 
> > > > > Could it be that 8xx also need a special treatment for kprobes?
> > > > 
> > > > Yes indeed, not for the same reasons but because 8xx ITLB-miss handler
> > > > doesn't expect any instruction TLB miss on kernel addresses other than 
> > > > the
> > > > ones from the module area:
> > > > 
> > > > Commit 4afb0be7ccda ("powerpc/8xx: Avoid testing for kernel space in 
> > > > ITLB
> > > > Miss.")
> > > > Commit 16a71c045186 ("powerpc/8xx: Preallocate execmem page tables")
> > > > Commit 33c527522f39 ("powerpc/8xx: Inconditionally use task PGDIR in 
> > > > ITLB
> > > > misses")
> > 
> > Right. I looked at the above series which added all of this support [1].
> > So looks like 603 might also have this related problem (other than the
> > per segment N bit issue), since it's iTLB miss handler only walks the
> > task's PGDIR, unlike it's dTLB handler which falls back to reloading the
> > missing kernel PGD entry from swapper_pg_dir. So what I meant to say is
> > that, similar to 8xx, a new PGD entry created in the kernel's vmalloc
> > area for an insn page does not get copied over after the task is
> > created (/PGDIR allocated), on 603 as well.
> > But since MODLUES_VADDR is defined for both BOOK3S/32 and 8xx, so I guess it
> > gets covered by the same change anyways.
> > 
> > > > So at the end as MODULES_VADDR/MODULES_END only exist on book3s/32 and 
> > > > 8xx I
> > > > think when they exist they should be used as follows:
> > > > 
> > > > diff --git a/arch/powerpc/mm/mem.c b/arch/powerpc/mm/mem.c
> > > > index b617b69452cd..856f4609c488 100644
> > > > --- a/arch/powerpc/mm/mem.c
> > > > +++ b/arch/powerpc/mm/mem.c
> > > > @@ -428,8 +428,13 @@ struct execmem_info __init 
> > > > *execmem_arch_setup(void)
> > > >                                 .fallback_end   = fallback_end,
> > > >                         },
> > > >                         [EXECMEM_KPROBES] = {
> > > > +#ifdef MODULES_VADDR
> > > > +                               .start  = MODULES_VADDR,
> > > > +                               .end    = MODULES_END,
> > > > +#else
> > > >                                 .start  = VMALLOC_START,
> > > >                                 .end    = VMALLOC_END,
> > > > +#endif
> > > 
> > > We already have #ifdef MODULES_VADDR a few lines above, I think the
> > > assignment of kprobes_start and kprobes_end could go there instead of if
> > > (IS_ENABLED(CONFIG_PPC_BOOK3S_32)).
> > > 
> > 
> > Right. So let me use the following diff then?
> > 
> > Please let me know if this diff looks good to you? If yes, then I can
> > send a v3.
> 
> My preference would go to something similar to 'start' and 'end', no init at
> declaration, kprobes_start and kprobes_end set to VMALLOC_START/VMALLOC_END
> inside the #else

I agree, it's more consistent this way.
 
> > Also - it will be good to know if patch-1 and patch-3 looks ok too,
> > before I spin a next revision?

Both make sense to me, feel free to add my

Acked-by: Mike Rapoport (Microsoft) <[email protected]>

-- 
Sincerely yours,
Mike.

Reply via email to