linux/arch/ia64/kernel
Tony Luck c0b5a64d93 [IA64] Change default PSR.ac from '1' to '0' (Fix erratum #237)
April 2014 Itanium processor specification update:

http://www.intel.com/content/www/us/en/processors/itanium/itanium-specification-update.html

describes this erratum:

=========================================================================
237. Under a complex set of conditions, store to load forwarding for a
sub 8-byte load may complete incorrectly

Problem: A load instruction may complete incorrectly when a code sequence
using 4-byte or smaller load and store operations to the same address
is executed in combination with specific timing of all the following
concurrent conditions: store to load forwarding, alignment checking
enabled, a mis-predicted branch, and complex cache utilization activity.

Implication: The affected sub 8-byte instruction may complete
incorrectly resulting in unpredictable system behavior. There is an
extremely low probability of exposure due to the significant number of
complex microarchitectural concurrent conditions required to encounter
the erratum.

Workaround: Set PSR.ac = 0 to completely avoid the erratum. Disabling
Hyper-Threading will significantly reduce exposure to the conditions
that contribute to encountering the erratum.

Status: See the Summary Table of Changes for the affected steppings.
=========================================================================

[Table of changes essentially lists all models from McKinley to Tukwila]

The PSR.ac bit controls whether the processor will always generate
an unaligned reference trap (0x5a00) for a misaligned data access
(when PSR.ac=1) or if it will let the access succeed when running
on a cpu that implements logic to handle some unaligned accesses.

Way back in 2008 in commit b704882e70
  [IA64] Rationalize kernel mode alignment checking
we made the decision to always enable strict checking. We were
already doing so in trap/interrupt context because the common
preamble code set this bit - but the rest of supervisor code
(and by inheritance user code) ran with PSR.ac=0.

We now reverse that decision and set PSR.ac=0 everywhere in the
kernel (also inherited by user processes). This will avoid the
erratum using the method described in the Itanium specification
update.  Net effect for users is that the processor will handle
unaligned access when it can (typically with a tiny performance
bubble in the pipeline ... but much less invasive than taking a
trap and having the OS perform the access).

Signed-off-by: Tony Luck <tony.luck@intel.com>
2014-04-16 10:20:34 -07:00
..
.gitignore
acpi-ext.c
acpi.c ACPI and power management updates for 3.14-rc1 2014-01-24 15:51:02 -08:00
asm-offsets.c ia64/xen: Remove Xen support for ia64 2013-12-10 16:11:07 -08:00
audit.c
brl_emu.c
crash.c
crash_dump.c
cyclone.c
dma-mapping.c
efi.c efi: x86: ia64: provide a generic efi_config_init() 2013-09-05 13:29:29 +01:00
efi_stub.S
elfcore.c switch elf_core_write_extra_data() to dump_emit() 2013-11-09 00:16:23 -05:00
entry.h
entry.S [IA64] Wire up new sched_setattr and sched_getattr syscalls 2014-01-28 09:52:53 -08:00
err_inject.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
esi.c
esi_stub.S
fsys.S Fix broken fsys_getppid() 2013-03-19 16:14:52 -07:00
fsyscall_gtod_data.h
ftrace.c
gate-data.S
gate.lds.S
gate.S
head.S [IA64] Change default PSR.ac from '1' to '0' (Fix erratum #237) 2014-04-16 10:20:34 -07:00
ia64_ksyms.c treewide: Put a space between #include and FILE 2012-06-28 11:44:36 +02:00
init_task.c
iosapic.c Fix kexec oops when iosapic was removed 2013-03-19 16:14:53 -07:00
irq.c Fix initialization of CMCI/CMCP interrupts 2013-04-02 09:37:06 -07:00
irq_ia64.c random: remove rand_initialize_irq() 2012-07-19 10:38:32 -04:00
irq_lsapic.c
ivt.S [IA64] Change default PSR.ac from '1' to '0' (Fix erratum #237) 2014-04-16 10:20:34 -07:00
jprobes.S
kprobes.c doc: typo on word accounting in kprobes.c in mutliple architectures 2013-10-14 15:46:39 +02:00
machine_kexec.c [IA64] kexec: Move the dereference below the NULL test 2012-09-10 14:31:23 -07:00
machvec.c
Makefile cpufreq: ia64: move cpufreq driver to drivers/cpufreq 2013-04-10 13:19:25 +02:00
Makefile.gate
mca.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
mca_asm.S
mca_drv.c Remove cast for kmalloc return value 2013-03-19 16:14:53 -07:00
mca_drv.h
mca_drv_asm.S
minstate.h cputime: Generic on-demand virtual cputime accounting 2013-01-27 19:23:27 +01:00
module.c
msi_ia64.c
nr-irqs.c ia64/xen: Remove Xen support for ia64 2013-12-10 16:11:07 -08:00
numa.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
pal.S
palinfo.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
paravirt.c
paravirt_inst.h ia64/xen: Remove Xen support for ia64 2013-12-10 16:11:07 -08:00
paravirt_patch.c
paravirt_patchlist.c
paravirt_patchlist.h ia64/xen: Remove Xen support for ia64 2013-12-10 16:11:07 -08:00
paravirtentry.S
patch.c
pci-dma.c [IA64] pci: Remove unused fallback_dev 2013-06-03 16:04:31 -07:00
pci-swiotlb.c
perfmon.c consolidate simple ->d_delete() instances 2013-11-15 22:04:17 -05:00
perfmon_default_smpl.c
perfmon_generic.h
perfmon_itanium.h
perfmon_mckinley.h
perfmon_montecito.h
process.c dump_stack: unify debug information printed by show_regs() 2013-04-30 17:04:02 -07:00
ptrace.c ia64: kill thread_matches(), unexport ptrace_check_attach() 2013-01-20 12:26:05 -08:00
relocate_kernel.S
sal.c
salinfo.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
setup.c DMI: Parse memory device (type 17) in SMBIOS 2013-10-23 10:10:12 -07:00
sigframe.h
signal.c constify copy_siginfo_to_user{,32}() 2013-11-09 00:16:29 -05:00
smp.c
smpboot.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
stacktrace.c
sys_ia64.c mm: use vm_unmapped_area() on ia64 architecture 2013-02-22 13:46:59 -08:00
time.c kvm: Prepare to add generic guest entry/exit callbacks 2013-01-27 20:35:40 +01:00
topology.c [IA64] Delete __cpuinit usage from all ia64 users 2013-06-24 15:44:27 -07:00
traps.c [IA64] Fix trap #45 handling 2013-06-03 15:47:42 -07:00
unaligned.c
uncached.c mm: fix GFP_THISNODE callers and clarify 2014-03-10 17:26:19 -07:00
unwind.c
unwind_decoder.c
unwind_i.h
vmlinux.lds.S ia64/xen: Remove Xen support for ia64 2013-12-10 16:11:07 -08:00