[Qemu-devel] [Bug 1587535] [NEW] Incorrect MAS1_TSIZE_SHIFT in ppce500_spin.c causes incorrectly sized TLB.

* [Qemu-devel] [Bug 1587535] [NEW] Incorrect MAS1_TSIZE_SHIFT in ppce500_spin.c causes incorrectly sized TLB.
@ 2016-05-31 15:32 Aaron Larson
  2016-06-30 13:20 ` [Qemu-devel] [Bug 1587535] " Aaron Larson
  2017-01-19 14:21 ` Thomas Huth
  0 siblings, 2 replies; 3+ messages in thread
From: Aaron Larson @ 2016-05-31 15:32 UTC (permalink / raw)
  To: qemu-devel

Public bug reported:

When e500 PPC is booted multi-core, the non-boot cores are started via
the spin table.  ppce500_spin.c:spin_kick() calls
mmubooke_create_initial_mapping() to allocate a 64MB TLB entry, but
the created TLB entry is only 256KB.

The root cause is that the function computing the size of the TLB
entry, namely booke206_page_size_to_tlb assumes MAS1.TSIZE as defined
by latter PPC cores, specifically n to the power of FOUR * 1KB.  The
result is then used by mmubooke_create_initial_mapping using
MAS1_TSIZE_SHIFT, but MAS1_TSIZE_SHIFT is defined assuming TLB entries
are n to the power of TWO * 1KB.  I.e., a difference of shift=7 or
shift=8.

Simply changing MAS1_TSIZE_SHIFT from 7 to 8 is not appropriate since
the macro is used elsewhere.

Removing the ">>1" from:

> static inline hwaddr booke206_page_size_to_tlb(uint64_t size)
> {
>     return ctz32(size >> 10) >> 1;

and adding an appropriate comment is what I used as a work around:

> static inline hwaddr booke206_page_size_to_tlb(uint64_t size)
> {
>     // resulting size is based on MAS1_TSIZE_SHIFT=7 TLB size.
>     return ctz32(size >> 10);

** Affects: qemu
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of qemu-
devel-ml, which is subscribed to QEMU.
https://bugs.launchpad.net/bugs/1587535

Title:
  Incorrect MAS1_TSIZE_SHIFT in ppce500_spin.c causes incorrectly sized
  TLB.

Status in QEMU:
  New

Bug description:
  When e500 PPC is booted multi-core, the non-boot cores are started via
  the spin table.  ppce500_spin.c:spin_kick() calls
  mmubooke_create_initial_mapping() to allocate a 64MB TLB entry, but
  the created TLB entry is only 256KB.

  The root cause is that the function computing the size of the TLB
  entry, namely booke206_page_size_to_tlb assumes MAS1.TSIZE as defined
  by latter PPC cores, specifically n to the power of FOUR * 1KB.  The
  result is then used by mmubooke_create_initial_mapping using
  MAS1_TSIZE_SHIFT, but MAS1_TSIZE_SHIFT is defined assuming TLB entries
  are n to the power of TWO * 1KB.  I.e., a difference of shift=7 or
  shift=8.

  Simply changing MAS1_TSIZE_SHIFT from 7 to 8 is not appropriate since
  the macro is used elsewhere.

  Removing the ">>1" from:

  > static inline hwaddr booke206_page_size_to_tlb(uint64_t size)
  > {
  >     return ctz32(size >> 10) >> 1;

  and adding an appropriate comment is what I used as a work around:

  > static inline hwaddr booke206_page_size_to_tlb(uint64_t size)
  > {
  >     // resulting size is based on MAS1_TSIZE_SHIFT=7 TLB size.
  >     return ctz32(size >> 10);

To manage notifications about this bug go to:
https://bugs.launchpad.net/qemu/+bug/1587535/+subscriptions

^ permalink raw reply	[flat|nested] 3+ messages in thread