Patch "powerpc/mm: Fixup tlbie vs mtpidr/mtlpidr ordering issue on POWER9" has been added to the 4.14-stable tree
gregkh at linuxfoundation.org
gregkh at linuxfoundation.org
Sat Nov 9 03:56:25 AEDT 2019
This is a note to let you know that I've just added the patch titled
powerpc/mm: Fixup tlbie vs mtpidr/mtlpidr ordering issue on POWER9
to the 4.14-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
powerpc-mm-fixup-tlbie-vs-mtpidr-mtlpidr-ordering-issue-on-power9.patch
and it can be found in the queue-4.14 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable at vger.kernel.org> know about it.
>From foo at baz Fri 08 Nov 2019 05:55:09 PM CET
From: Sandipan Das <sandipan at linux.ibm.com>
Date: Thu, 17 Oct 2019 13:35:03 +0530
Subject: powerpc/mm: Fixup tlbie vs mtpidr/mtlpidr ordering issue on POWER9
To: gregkh at linuxfoundation.org
Cc: stable at vger.kernel.org, aneesh.kumar at linux.ibm.com, mpe at ellerman.id.au, linuxppc-dev at lists.ozlabs.org
Message-ID: <20191017080505.8348-4-sandipan at linux.ibm.com>
From: "Aneesh Kumar K.V" <aneesh.kumar at linux.ibm.com>
commit 047e6575aec71d75b765c22111820c4776cd1c43 upstream.
On POWER9, under some circumstances, a broadcast TLB invalidation will
fail to invalidate the ERAT cache on some threads when there are
parallel mtpidr/mtlpidr happening on other threads of the same core.
This can cause stores to continue to go to a page after it's unmapped.
The workaround is to force an ERAT flush using PID=0 or LPID=0 tlbie
flush. This additional TLB flush will cause the ERAT cache
invalidation. Since we are using PID=0 or LPID=0, we don't get
filtered out by the TLB snoop filtering logic.
We need to still follow this up with another tlbie to take care of
store vs tlbie ordering issue explained in commit:
a5d4b5891c2f ("powerpc/mm: Fixup tlbie vs store ordering issue on
POWER9"). The presence of ERAT cache implies we can still get new
stores and they may miss store queue marking flush.
Cc: stable at vger.kernel.org # v4.14
Signed-off-by: Aneesh Kumar K.V <aneesh.kumar at linux.ibm.com>
Signed-off-by: Michael Ellerman <mpe at ellerman.id.au>
Link: https://lore.kernel.org/r/20190924035254.24612-3-aneesh.kumar@linux.ibm.com
[sandipan: Backported to v4.14]
Signed-off-by: Sandipan Das <sandipan at linux.ibm.com>
Signed-off-by: Greg Kroah-Hartman <gregkh at linuxfoundation.org>
---
arch/powerpc/include/asm/cputable.h | 3 +
arch/powerpc/kernel/dt_cpu_ftrs.c | 2 +
arch/powerpc/kvm/book3s_hv_rm_mmu.c | 42 +++++++++++++++++------
arch/powerpc/mm/hash_native_64.c | 28 +++++++++++++--
arch/powerpc/mm/tlb-radix.c | 65 ++++++++++++++++++++++++++++++------
5 files changed, 116 insertions(+), 24 deletions(-)
--- a/arch/powerpc/include/asm/cputable.h
+++ b/arch/powerpc/include/asm/cputable.h
@@ -217,6 +217,7 @@ enum {
#define CPU_FTR_PMAO_BUG LONG_ASM_CONST(0x1000000000000000)
#define CPU_FTR_P9_TLBIE_STQ_BUG LONG_ASM_CONST(0x0000400000000000)
#define CPU_FTR_POWER9_DD1 LONG_ASM_CONST(0x4000000000000000)
+#define CPU_FTR_P9_TLBIE_ERAT_BUG LONG_ASM_CONST(0x0001000000000000)
#ifndef __ASSEMBLY__
@@ -477,7 +478,7 @@ enum {
CPU_FTR_CFAR | CPU_FTR_HVMODE | CPU_FTR_VMX_COPY | \
CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_DAWR | \
CPU_FTR_ARCH_207S | CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | \
- CPU_FTR_P9_TLBIE_STQ_BUG)
+ CPU_FTR_P9_TLBIE_STQ_BUG | CPU_FTR_P9_TLBIE_ERAT_BUG)
#define CPU_FTRS_POWER9_DD1 ((CPU_FTRS_POWER9 | CPU_FTR_POWER9_DD1) & \
(~CPU_FTR_SAO))
#define CPU_FTRS_CELL (CPU_FTR_USE_TB | CPU_FTR_LWSYNC | \
--- a/arch/powerpc/kernel/dt_cpu_ftrs.c
+++ b/arch/powerpc/kernel/dt_cpu_ftrs.c
@@ -756,6 +756,8 @@ static __init void update_tlbie_feature_
WARN_ONCE(1, "Unknown PVR");
cur_cpu_spec->cpu_features |= CPU_FTR_P9_TLBIE_STQ_BUG;
}
+
+ cur_cpu_spec->cpu_features |= CPU_FTR_P9_TLBIE_ERAT_BUG;
}
}
--- a/arch/powerpc/kvm/book3s_hv_rm_mmu.c
+++ b/arch/powerpc/kvm/book3s_hv_rm_mmu.c
@@ -429,6 +429,37 @@ static inline int try_lock_tlbie(unsigne
return old == 0;
}
+static inline void fixup_tlbie_lpid(unsigned long rb_value, unsigned long lpid)
+{
+
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_ERAT_BUG)) {
+ /* Radix flush for a hash guest */
+
+ unsigned long rb,rs,prs,r,ric;
+
+ rb = PPC_BIT(52); /* IS = 2 */
+ rs = 0; /* lpid = 0 */
+ prs = 0; /* partition scoped */
+ r = 1; /* radix format */
+ ric = 0; /* RIC_FLSUH_TLB */
+
+ /*
+ * Need the extra ptesync to make sure we don't
+ * re-order the tlbie
+ */
+ asm volatile("ptesync": : :"memory");
+ asm volatile(PPC_TLBIE_5(%0, %4, %3, %2, %1)
+ : : "r"(rb), "i"(r), "i"(prs),
+ "i"(ric), "r"(rs) : "memory");
+ }
+
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_STQ_BUG)) {
+ asm volatile("ptesync": : :"memory");
+ asm volatile(PPC_TLBIE_5(%0,%1,0,0,0) : :
+ "r" (rb_value), "r" (lpid));
+ }
+}
+
static void do_tlbies(struct kvm *kvm, unsigned long *rbvalues,
long npages, int global, bool need_sync)
{
@@ -449,16 +480,7 @@ static void do_tlbies(struct kvm *kvm, u
"r" (rbvalues[i]), "r" (kvm->arch.lpid));
}
- if (cpu_has_feature(CPU_FTR_P9_TLBIE_STQ_BUG)) {
- /*
- * Need the extra ptesync to make sure we don't
- * re-order the tlbie
- */
- asm volatile("ptesync": : :"memory");
- asm volatile(PPC_TLBIE_5(%0,%1,0,0,0) : :
- "r" (rbvalues[0]), "r" (kvm->arch.lpid));
- }
-
+ fixup_tlbie_lpid(rbvalues[i - 1], kvm->arch.lpid);
asm volatile("eieio; tlbsync; ptesync" : : : "memory");
kvm->arch.tlbie_lock = 0;
} else {
--- a/arch/powerpc/mm/hash_native_64.c
+++ b/arch/powerpc/mm/hash_native_64.c
@@ -104,8 +104,30 @@ static inline unsigned long ___tlbie(un
return va;
}
-static inline void fixup_tlbie(unsigned long vpn, int psize, int apsize, int ssize)
+static inline void fixup_tlbie_vpn(unsigned long vpn, int psize,
+ int apsize, int ssize)
{
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_ERAT_BUG)) {
+ /* Radix flush for a hash guest */
+
+ unsigned long rb,rs,prs,r,ric;
+
+ rb = PPC_BIT(52); /* IS = 2 */
+ rs = 0; /* lpid = 0 */
+ prs = 0; /* partition scoped */
+ r = 1; /* radix format */
+ ric = 0; /* RIC_FLSUH_TLB */
+
+ /*
+ * Need the extra ptesync to make sure we don't
+ * re-order the tlbie
+ */
+ asm volatile("ptesync": : :"memory");
+ asm volatile(PPC_TLBIE_5(%0, %4, %3, %2, %1)
+ : : "r"(rb), "i"(r), "i"(prs),
+ "i"(ric), "r"(rs) : "memory");
+ }
+
if (cpu_has_feature(CPU_FTR_P9_TLBIE_STQ_BUG)) {
/* Need the extra ptesync to ensure we don't reorder tlbie*/
asm volatile("ptesync": : :"memory");
@@ -190,7 +212,7 @@ static inline void tlbie(unsigned long v
asm volatile("ptesync": : :"memory");
} else {
__tlbie(vpn, psize, apsize, ssize);
- fixup_tlbie(vpn, psize, apsize, ssize);
+ fixup_tlbie_vpn(vpn, psize, apsize, ssize);
asm volatile("eieio; tlbsync; ptesync": : :"memory");
}
if (lock_tlbie && !use_local)
@@ -759,7 +781,7 @@ static void native_flush_hash_range(unsi
/*
* Just do one more with the last used values.
*/
- fixup_tlbie(vpn, psize, psize, ssize);
+ fixup_tlbie_vpn(vpn, psize, psize, ssize);
asm volatile("eieio; tlbsync; ptesync":::"memory");
if (lock_tlbie)
--- a/arch/powerpc/mm/tlb-radix.c
+++ b/arch/powerpc/mm/tlb-radix.c
@@ -39,14 +39,18 @@ static inline void __tlbie_va(unsigned l
trace_tlbie(0, 0, rb, rs, ric, prs, r);
}
-static inline void fixup_tlbie(void)
+
+static inline void fixup_tlbie_va(unsigned long va, unsigned long pid,
+ unsigned long ap)
{
- unsigned long pid = 0;
- unsigned long va = ((1UL << 52) - 1);
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_ERAT_BUG)) {
+ asm volatile("ptesync": : :"memory");
+ __tlbie_va(va, 0, ap, RIC_FLUSH_TLB);
+ }
if (cpu_has_feature(CPU_FTR_P9_TLBIE_STQ_BUG)) {
asm volatile("ptesync": : :"memory");
- __tlbie_va(va, pid, mmu_get_ap(MMU_PAGE_64K), RIC_FLUSH_TLB);
+ __tlbie_va(va, pid, ap, RIC_FLUSH_TLB);
}
}
@@ -95,23 +99,64 @@ static inline void _tlbiel_pid(unsigned
asm volatile(PPC_INVALIDATE_ERAT "; isync" : : :"memory");
}
-static inline void _tlbie_pid(unsigned long pid, unsigned long ric)
+static inline void __tlbie_pid(unsigned long pid, unsigned long ric)
{
unsigned long rb,rs,prs,r;
rb = PPC_BIT(53); /* IS = 1 */
rs = pid << PPC_BITLSHIFT(31);
prs = 1; /* process scoped */
- r = 1; /* raidx format */
+ r = 1; /* radix format */
- asm volatile("ptesync": : :"memory");
asm volatile(PPC_TLBIE_5(%0, %4, %3, %2, %1)
: : "r"(rb), "i"(r), "i"(prs), "i"(ric), "r"(rs) : "memory");
- fixup_tlbie();
- asm volatile("eieio; tlbsync; ptesync": : :"memory");
trace_tlbie(0, 0, rb, rs, ric, prs, r);
}
+static inline void fixup_tlbie_pid(unsigned long pid)
+{
+ /*
+ * We can use any address for the invalidation, pick one which is
+ * probably unused as an optimisation.
+ */
+ unsigned long va = ((1UL << 52) - 1);
+
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_ERAT_BUG)) {
+ asm volatile("ptesync": : :"memory");
+ __tlbie_pid(0, RIC_FLUSH_TLB);
+ }
+
+ if (cpu_has_feature(CPU_FTR_P9_TLBIE_STQ_BUG)) {
+ asm volatile("ptesync": : :"memory");
+ __tlbie_va(va, pid, mmu_get_ap(MMU_PAGE_64K), RIC_FLUSH_TLB);
+ }
+}
+
+static inline void _tlbie_pid(unsigned long pid, unsigned long ric)
+{
+ asm volatile("ptesync": : :"memory");
+
+ /*
+ * Workaround the fact that the "ric" argument to __tlbie_pid
+ * must be a compile-time contraint to match the "i" constraint
+ * in the asm statement.
+ */
+ switch (ric) {
+ case RIC_FLUSH_TLB:
+ __tlbie_pid(pid, RIC_FLUSH_TLB);
+ fixup_tlbie_pid(pid);
+ break;
+ case RIC_FLUSH_PWC:
+ __tlbie_pid(pid, RIC_FLUSH_PWC);
+ break;
+ case RIC_FLUSH_ALL:
+ default:
+ __tlbie_pid(pid, RIC_FLUSH_ALL);
+ fixup_tlbie_pid(pid);
+ }
+ asm volatile("eieio; tlbsync; ptesync": : :"memory");
+}
+
static inline void _tlbiel_va(unsigned long va, unsigned long pid,
unsigned long ap, unsigned long ric)
{
@@ -135,7 +180,7 @@ static inline void _tlbie_va(unsigned lo
{
asm volatile("ptesync": : :"memory");
__tlbie_va(va, pid, ap, ric);
- fixup_tlbie();
+ fixup_tlbie_va(va, pid, ap);
asm volatile("eieio; tlbsync; ptesync": : :"memory");
}
Patches currently in stable-queue which might be from sandipan at linux.ibm.com are
queue-4.14/powerpc-book3s64-mm-don-t-do-tlbie-fixup-for-some-hardware-revisions.patch
queue-4.14/selftests-powerpc-add-test-case-for-tlbie-vs-mtpidr-ordering-issue.patch
queue-4.14/selftests-powerpc-fix-compile-error-on-tlbie_test-due-to-newer-gcc.patch
queue-4.14/powerpc-mm-fixup-tlbie-vs-mtpidr-mtlpidr-ordering-issue-on-power9.patch
queue-4.14/powerpc-mm-fixup-tlbie-vs-store-ordering-issue-on-power9.patch
queue-4.14/powerpc-book3s64-radix-rename-cpu_ftr_p9_tlbie_bug-feature-flag.patch
More information about the Linuxppc-dev
mailing list