Re: [PATCH v6 4/4] powerpc/64s: Initialize and use a temporary mm for patching on Radix

[Jordan Niethe <jniethe5@gmail.com>]

On Thu, Sep 16, 2021 at 10:38 AM Christopher M. Riedl
<cmr@bluescreens.de> wrote:
>
> On Sat Sep 11, 2021 at 4:14 AM CDT, Jordan Niethe wrote:
> > On Sat, Sep 11, 2021 at 12:39 PM Christopher M. Riedl
> > <cmr@bluescreens.de> wrote:
> > >
> > > When code patching a STRICT_KERNEL_RWX kernel the page containing the
> > > address to be patched is temporarily mapped as writeable. Currently, a
> > > per-cpu vmalloc patch area is used for this purpose. While the patch
> > > area is per-cpu, the temporary page mapping is inserted into the kernel
> > > page tables for the duration of patching. The mapping is exposed to CPUs
> > > other than the patching CPU - this is undesirable from a hardening
> > > perspective. Use a temporary mm instead which keeps the mapping local to
> > > the CPU doing the patching.
> > >
> > > Use the `poking_init` init hook to prepare a temporary mm and patching
> > > address. Initialize the temporary mm by copying the init mm. Choose a
> > > randomized patching address inside the temporary mm userspace address
> > > space. The patching address is randomized between PAGE_SIZE and
> > > DEFAULT_MAP_WINDOW-PAGE_SIZE.
> > >
> > > Bits of entropy with 64K page size on BOOK3S_64:
> > >
> > >         bits of entropy = log2(DEFAULT_MAP_WINDOW_USER64 / PAGE_SIZE)
> > >
> > >         PAGE_SIZE=64K, DEFAULT_MAP_WINDOW_USER64=128TB
> > >         bits of entropy = log2(128TB / 64K)
> > >         bits of entropy = 31
> > >
> > > The upper limit is DEFAULT_MAP_WINDOW due to how the Book3s64 Hash MMU
> > > operates - by default the space above DEFAULT_MAP_WINDOW is not
> > > available. Currently the Hash MMU does not use a temporary mm so
> > > technically this upper limit isn't necessary; however, a larger
> > > randomization range does not further "harden" this overall approach and
> > > future work may introduce patching with a temporary mm on Hash as well.
> > >
> > > Randomization occurs only once during initialization at boot for each
> > > possible CPU in the system.
> > >
> > > Introduce two new functions, map_patch_mm() and unmap_patch_mm(), to
> > > respectively create and remove the temporary mapping with write
> > > permissions at patching_addr. Map the page with PAGE_KERNEL to set
> > > EAA[0] for the PTE which ignores the AMR (so no need to unlock/lock
> > > KUAP) according to PowerISA v3.0b Figure 35 on Radix.
> > >
> > > Based on x86 implementation:
> > >
> > > commit 4fc19708b165
> > > ("x86/alternatives: Initialize temporary mm for patching")
> > >
> > > and:
> > >
> > > commit b3fd8e83ada0
> > > ("x86/alternatives: Use temporary mm for text poking")
> > >
> > > Signed-off-by: Christopher M. Riedl <cmr@bluescreens.de>
> > >
> > > ---
> > >
> > > v6:  * Small clean-ups (naming, formatting, style, etc).
> > >      * Call stop_using_temporary_mm() before pte_unmap_unlock() after
> > >        patching.
> > >      * Replace BUG_ON()s in poking_init() w/ WARN_ON()s.
> > >
> > > v5:  * Only support Book3s64 Radix MMU for now.
> > >      * Use a per-cpu datastructure to hold the patching_addr and
> > >        patching_mm to avoid the need for a synchronization lock/mutex.
> > >
> > > v4:  * In the previous series this was two separate patches: one to init
> > >        the temporary mm in poking_init() (unused in powerpc at the time)
> > >        and the other to use it for patching (which removed all the
> > >        per-cpu vmalloc code). Now that we use poking_init() in the
> > >        existing per-cpu vmalloc approach, that separation doesn't work
> > >        as nicely anymore so I just merged the two patches into one.
> > >      * Preload the SLB entry and hash the page for the patching_addr
> > >        when using Hash on book3s64 to avoid taking an SLB and Hash fault
> > >        during patching. The previous implementation was a hack which
> > >        changed current->mm to allow the SLB and Hash fault handlers to
> > >        work with the temporary mm since both of those code-paths always
> > >        assume mm == current->mm.
> > >      * Also (hmm - seeing a trend here) with the book3s64 Hash MMU we
> > >        have to manage the mm->context.active_cpus counter and mm cpumask
> > >        since they determine (via mm_is_thread_local()) if the TLB flush
> > >        in pte_clear() is local or not - it should always be local when
> > >        we're using the temporary mm. On book3s64's Radix MMU we can
> > >        just call local_flush_tlb_mm().
> > >      * Use HPTE_USE_KERNEL_KEY on Hash to avoid costly lock/unlock of
> > >        KUAP.
> > > ---
> > >  arch/powerpc/lib/code-patching.c | 119 +++++++++++++++++++++++++++++--
> > >  1 file changed, 112 insertions(+), 7 deletions(-)
> > >
> > > diff --git a/arch/powerpc/lib/code-patching.c b/arch/powerpc/lib/code-patching.c
> > > index e802e42c2789..af8e2a02a9dd 100644
> > > --- a/arch/powerpc/lib/code-patching.c
> > > +++ b/arch/powerpc/lib/code-patching.c
> > > @@ -11,6 +11,7 @@
> > >  #include <linux/cpuhotplug.h>
> > >  #include <linux/slab.h>
> > >  #include <linux/uaccess.h>
> > > +#include <linux/random.h>
> > >
> > >  #include <asm/tlbflush.h>
> > >  #include <asm/page.h>
> > > @@ -103,6 +104,7 @@ static inline void stop_using_temporary_mm(struct temp_mm *temp_mm)
> > >
> > >  static DEFINE_PER_CPU(struct vm_struct *, text_poke_area);
> > >  static DEFINE_PER_CPU(unsigned long, cpu_patching_addr);
> > > +static DEFINE_PER_CPU(struct mm_struct *, cpu_patching_mm);
> > >
> > >  static int text_area_cpu_up(unsigned int cpu)
> > >  {
> > > @@ -126,8 +128,48 @@ static int text_area_cpu_down(unsigned int cpu)
> > >         return 0;
> > >  }
> > >
> > > +static __always_inline void __poking_init_temp_mm(void)
> > > +{
> > > +       int cpu;
> > > +       spinlock_t *ptl; /* for protecting pte table */
> >
> > ptl is just used so we don't have to open code allocating a pte in
> > patching_mm isn't it?
>
> Yup - I think that comment was a copy-pasta... I'll improve it.
>
> >
> > > +       pte_t *ptep;
> > > +       struct mm_struct *patching_mm;
> > > +       unsigned long patching_addr;
> > > +
> > > +       for_each_possible_cpu(cpu) {
> > > +               patching_mm = copy_init_mm();
> > > +               WARN_ON(!patching_mm);
> >
> > Would it be okay to just let the mmu handle null pointer dereferences?
>
> In general I think yes; however, the NULL dereference wouldn't occur
> until later during actual patching so I thought an early WARN here is
> appropriate.
>
> >
> > > +               per_cpu(cpu_patching_mm, cpu) = patching_mm;
> > > +
> > > +               /*
> > > +                * Choose a randomized, page-aligned address from the range:
> > > +                * [PAGE_SIZE, DEFAULT_MAP_WINDOW - PAGE_SIZE] The lower
> > > +                * address bound is PAGE_SIZE to avoid the zero-page.  The
> > > +                * upper address bound is DEFAULT_MAP_WINDOW - PAGE_SIZE to
> > > +                * stay under DEFAULT_MAP_WINDOW with the Book3s64 Hash MMU.
> > > +                */
> > > +               patching_addr = PAGE_SIZE + ((get_random_long() & PAGE_MASK)
> > > +                               % (DEFAULT_MAP_WINDOW - 2 * PAGE_SIZE));
> > > +               per_cpu(cpu_patching_addr, cpu) = patching_addr;
> >
> > On x86 the randomization depends on CONFIG_RANDOMIZE_BASE. Should it
> > be controllable here too?
>
> IIRC CONFIG_RANDOMIZE_BASE is for KASLR which IMO doesn't really have
> much to do with this.
>
> >
> > > +
> > > +               /*
> > > +                * PTE allocation uses GFP_KERNEL which means we need to
> > > +                * pre-allocate the PTE here because we cannot do the
> > > +                * allocation during patching when IRQs are disabled.
> > > +                */
> > > +               ptep = get_locked_pte(patching_mm, patching_addr, &ptl);
> > > +               WARN_ON(!ptep);
> > > +               pte_unmap_unlock(ptep, ptl);
> > > +       }
> > > +}
> > > +
> > >  void __init poking_init(void)
> > >  {
> > > +       if (radix_enabled()) {
> > > +               __poking_init_temp_mm();
> >
> > Should this also be done with cpuhp_setup_state()?
>
> I think I prefer doing the setup ahead of time during boot.

It does lose the ability to free up memory after a cpu is hot
unplugged but I'm not sure if that's a big problem.

>
> >
> > > +               return;
> > > +       }
> > > +
> > >         WARN_ON(cpuhp_setup_state(CPUHP_AP_ONLINE_DYN,
> > >                 "powerpc/text_poke:online", text_area_cpu_up,
> > >                 text_area_cpu_down) < 0);
> > > @@ -197,30 +239,93 @@ static inline int unmap_patch_area(void)
> > >         return 0;
> > >  }
> > >
> > > +struct patch_mapping {
> > > +       spinlock_t *ptl; /* for protecting pte table */
> > > +       pte_t *ptep;
> > > +       struct temp_mm temp_mm;
> > > +};
> > > +
> > > +/*
> > > + * This can be called for kernel text or a module.
> > > + */
> > > +static int map_patch_mm(const void *addr, struct patch_mapping *patch_mapping)
> > > +{
> > > +       struct page *page;
> > > +       struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
> > > +       unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
> > > +
> > > +       if (is_vmalloc_or_module_addr(addr))
> > > +               page = vmalloc_to_page(addr);
> > > +       else
> > > +               page = virt_to_page(addr);
> > > +
> > > +       patch_mapping->ptep = get_locked_pte(patching_mm, patching_addr,
> > > +                                            &patch_mapping->ptl);
> > > +       if (unlikely(!patch_mapping->ptep)) {
> > > +               pr_warn("map patch: failed to allocate pte for patching\n");
> > > +               return -1;
> > > +       }
> > > +
> > > +       set_pte_at(patching_mm, patching_addr, patch_mapping->ptep,
> > > +                  pte_mkdirty(mk_pte(page, PAGE_KERNEL)));
> > > +
> > > +       init_temp_mm(&patch_mapping->temp_mm, patching_mm);
> > > +       start_using_temporary_mm(&patch_mapping->temp_mm);
> > > +
> > > +       return 0;
> > > +}
> > > +
> > > +static int unmap_patch_mm(struct patch_mapping *patch_mapping)
> > > +{
> > > +       struct mm_struct *patching_mm = __this_cpu_read(cpu_patching_mm);
> > > +       unsigned long patching_addr = __this_cpu_read(cpu_patching_addr);
> > > +
> > > +       pte_clear(patching_mm, patching_addr, patch_mapping->ptep);
> > > +
> > > +       local_flush_tlb_mm(patching_mm);
> > > +       stop_using_temporary_mm(&patch_mapping->temp_mm);
> > > +
> > > +       pte_unmap_unlock(patch_mapping->ptep, patch_mapping->ptl);
> > > +
> > > +       return 0;
> > > +}
> > > +
> > >  static int do_patch_instruction(u32 *addr, struct ppc_inst instr)
> > >  {
> > >         int err, rc = 0;
> > >         u32 *patch_addr = NULL;
> > >         unsigned long flags;
> > > +       struct patch_mapping patch_mapping;
> > >
> > >         /*
> > > -        * During early early boot patch_instruction is called
> > > -        * when text_poke_area is not ready, but we still need
> > > -        * to allow patching. We just do the plain old patching
> > > +        * During early early boot patch_instruction is called when the
> > > +        * patching_mm/text_poke_area is not ready, but we still need to allow
> > > +        * patching. We just do the plain old patching.
> > >          */
> > > -       if (!this_cpu_read(text_poke_area))
> > > -               return raw_patch_instruction(addr, instr);
> > > +       if (radix_enabled()) {
> > > +               if (!this_cpu_read(cpu_patching_mm))
> > > +                       return raw_patch_instruction(addr, instr);
> > > +       } else {
> > > +               if (!this_cpu_read(text_poke_area))
> > > +                       return raw_patch_instruction(addr, instr);
> > > +       }
> >
> > Would testing cpu_patching_addr handler both of these cases?
> >
> > Then I think it might be clearer to do something like this:
> > if (radix_enabled()) {
> > return patch_instruction_mm(addr, instr);
> > }
> >
> > patch_instruction_mm() would combine map_patch_mm(), then patching and
> > unmap_patch_mm() into one function.
> >
> > IMO, a bit of code duplication would be cleaner than checking multiple
> > times for radix_enabled() and having struct patch_mapping especially
> > for maintaining state.
>
> Hmm, I think it's a good idea - I'll give it a go for the next version.
> Thanks for the suggestion!
>
> >
> > >
> > >         local_irq_save(flags);
> > >
> > > -       err = map_patch_area(addr);
> > > +       if (radix_enabled())
> > > +               err = map_patch_mm(addr, &patch_mapping);
> > > +       else
> > > +               err = map_patch_area(addr);
> > >         if (err)
> > >                 goto out;
> > >
> > >         patch_addr = (u32 *)(__this_cpu_read(cpu_patching_addr) | offset_in_page(addr));
> > >         rc = __patch_instruction(addr, instr, patch_addr);
> > >
> > > -       err = unmap_patch_area();
> > > +       if (radix_enabled())
> > > +               err = unmap_patch_mm(&patch_mapping);
> > > +       else
> > > +               err = unmap_patch_area();
> > >
> > >  out:
> > >         local_irq_restore(flags);
> > > --
> > > 2.32.0
> > >
> > Thanks,
> > Jordan
>