powerpc/64s/hash: Fix hash_preload running with interrupts enabled

Commit 2f92447f9f96 ("powerpc/book3s64/hash: Use the pte_t address from the
caller") removed the local_irq_disable from hash_preload, but it was
required for more than just the page table walk: the hash pte busy bit is
effectively a lock which may be taken in interrupt context, and the local
update flag test must not be preempted before it's used.

This solves apparent lockups with perf interrupting __hash_page_64K. If
get_perf_callchain then also takes a hash fault on the same page while it
is already locked, it will loop forever taking hash faults, which looks like
this:

  cpu 0x49e: Vector: 100 (System Reset) at [c00000001a4f7d70]
      pc: c000000000072dc8: hash_page_mm+0x8/0x800
      lr: c00000000000c5a4: do_hash_page+0x24/0x38
      sp: c0002ac1cc69ac70
     msr: 8000000000081033
    current = 0xc0002ac1cc602e00
    paca    = 0xc00000001de1f280   irqmask: 0x03   irq_happened: 0x01
      pid   = 20118, comm = pread2_processe
  Linux version 5.8.0-rc6-00345-g1fad14f18bc6
  49e:mon> t
  [c0002ac1cc69ac70] c00000000000c5a4 do_hash_page+0x24/0x38 (unreliable)
  --- Exception: 300 (Data Access) at c00000000008fa60 __copy_tofrom_user_power7+0x20c/0x7ac
  [link register   ] c000000000335d10 copy_from_user_nofault+0xf0/0x150
  [c0002ac1cc69af70] c00032bf9fa3c880 (unreliable)
  [c0002ac1cc69afa0] c000000000109df0 read_user_stack_64+0x70/0xf0
  [c0002ac1cc69afd0] c000000000109fcc perf_callchain_user_64+0x15c/0x410
  [c0002ac1cc69b060] c000000000109c00 perf_callchain_user+0x20/0x40
  [c0002ac1cc69b080] c00000000031c6cc get_perf_callchain+0x25c/0x360
  [c0002ac1cc69b120] c000000000316b50 perf_callchain+0x70/0xa0
  [c0002ac1cc69b140] c000000000316ddc perf_prepare_sample+0x25c/0x790
  [c0002ac1cc69b1a0] c000000000317350 perf_event_output_forward+0x40/0xb0
  [c0002ac1cc69b220] c000000000306138 __perf_event_overflow+0x88/0x1a0
  [c0002ac1cc69b270] c00000000010cf70 record_and_restart+0x230/0x750
  [c0002ac1cc69b620] c00000000010d69c perf_event_interrupt+0x20c/0x510
  [c0002ac1cc69b730] c000000000027d9c performance_monitor_exception+0x4c/0x60
  [c0002ac1cc69b750] c00000000000b2f8 performance_monitor_common_virt+0x1b8/0x1c0
  --- Exception: f00 (Performance Monitor) at c0000000000cb5b0 pSeries_lpar_hpte_insert+0x0/0x160
  [link register   ] c0000000000846f0 __hash_page_64K+0x210/0x540
  [c0002ac1cc69ba50] 0000000000000000 (unreliable)
  [c0002ac1cc69bb00] c000000000073ae0 update_mmu_cache+0x390/0x3a0
  [c0002ac1cc69bb70] c00000000037f024 wp_page_copy+0x364/0xce0
  [c0002ac1cc69bc20] c00000000038272c do_wp_page+0xdc/0xa60
  [c0002ac1cc69bc70] c0000000003857bc handle_mm_fault+0xb9c/0x1b60
  [c0002ac1cc69bd50] c00000000006c434 __do_page_fault+0x314/0xc90
  [c0002ac1cc69be20] c00000000000c5c8 handle_page_fault+0x10/0x2c
  --- Exception: 300 (Data Access) at 00007fff8c861fe8
  SP (7ffff6b19660) is in userspace

Fixes: 2f92447f9f96 ("powerpc/book3s64/hash: Use the pte_t address from the caller")
Reported-by: Athira Rajeev <atrajeev@linux.vnet.ibm.com>
Reported-by: Anton Blanchard <anton@ozlabs.org>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Aneesh Kumar K.V <aneesh.kumar@linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20200727060947.10060-1-npiggin@gmail.com

1 files changed, 25 insertions, 0 deletions

diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 468169e33c86..9b9f92ad0e7a 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -1559,6 +1559,7 @@ static void hash_preload(struct mm_struct *mm, pte_t *ptep, unsigned long ea,
 	pgd_t *pgdir;
 	int rc, ssize, update_flags = 0;
 	unsigned long access = _PAGE_PRESENT | _PAGE_READ | (is_exec ? _PAGE_EXEC : 0);
+	unsigned long flags;
 
 	BUG_ON(get_region_id(ea) != USER_REGION_ID);
 
@@ -1592,6 +1593,28 @@ static void hash_preload(struct mm_struct *mm, pte_t *ptep, unsigned long ea,
 		return;
 #endif /* CONFIG_PPC_64K_PAGES */
 
+	/*
+	 * __hash_page_* must run with interrupts off, as it sets the
+	 * H_PAGE_BUSY bit. It's possible for perf interrupts to hit at any
+	 * time and may take a hash fault reading the user stack, see
+	 * read_user_stack_slow() in the powerpc/perf code.
+	 *
+	 * If that takes a hash fault on the same page as we lock here, it
+	 * will bail out when seeing H_PAGE_BUSY set, and retry the access
+	 * leading to an infinite loop.
+	 *
+	 * Disabling interrupts here does not prevent perf interrupts, but it
+	 * will prevent them taking hash faults (see the NMI test in
+	 * do_hash_page), then read_user_stack's copy_from_user_nofault will
+	 * fail and perf will fall back to read_user_stack_slow(), which
+	 * walks the Linux page tables.
+	 *
+	 * Interrupts must also be off for the duration of the
+	 * mm_is_thread_local test and update, to prevent preempt running the
+	 * mm on another CPU (XXX: this may be racy vs kthread_use_mm).
+	 */
+	local_irq_save(flags);
+
 	/* Is that local to this CPU ? */
 	if (mm_is_thread_local(mm))
 		update_flags |= HPTE_LOCAL_UPDATE;
@@ -1614,6 +1637,8 @@ static void hash_preload(struct mm_struct *mm, pte_t *ptep, unsigned long ea,
 				   mm_ctx_user_psize(&mm->context),
 				   mm_ctx_user_psize(&mm->context),
 				   pte_val(*ptep));
+
+	local_irq_restore(flags);
 }
 
 /*