[PATCH v3 0/3] Fix oops in shared-processor spinlocks

Fixes an oops when calling the shared-processor spinlock implementation
from a non-SP LPAR. Also take this opportunity to refactor
SHARED_PROCESSOR a bit.

Reference:  https://github.com/linuxppc/issues/issues/229

Changes since v2:
 - Directly call splpar_*_yield() to avoid duplicate call to
   is_shared_processor() in some cases

Changes since v1:
 - Improve comment wording to make it clear why the BOOK3S #ifdef is
   required in is_shared_processor() in spinlock.h
 - Replace empty #define of splpar_*_yield() with actual functions with
   empty bodies

Christopher M. Riedl (3):
  powerpc/spinlocks: Refactor SHARED_PROCESSOR
  powerpc/spinlocks: Rename SPLPAR-only spinlocks
  powerpc/spinlocks: Fix oops in shared-processor spinlocks

 arch/powerpc/include/asm/spinlock.h | 62 +++++++++++++++++++++--------
 arch/powerpc/lib/locks.c            |  6 +--
 2 files changed, 48 insertions(+), 20 deletions(-)

-- 
2.22.0

Determining if a processor is in shared processor mode is not a constant
so don't hide it behind a #define.

Signed-off-by: Christopher M. Riedl <redacted>
Reviewed-by: Andrew Donnellan <redacted>
---
 arch/powerpc/include/asm/spinlock.h | 24 ++++++++++++++++++------
 1 file changed, 18 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/include/asm/spinlock.h b/arch/powerpc/include/asm/spinlock.h
index a47f827bc5f1..dc5fcea1f006 100644
--- a/arch/powerpc/include/asm/spinlock.h
+++ b/arch/powerpc/include/asm/spinlock.h

@@ -101,15 +101,27 @@ static inline int arch_spin_trylock(arch_spinlock_t *lock)
 
 #if defined(CONFIG_PPC_SPLPAR)
 /* We only yield to the hypervisor if we are in shared processor mode */
-#define SHARED_PROCESSOR (lppaca_shared_proc(local_paca->lppaca_ptr))
 extern void __spin_yield(arch_spinlock_t *lock);
 extern void __rw_yield(arch_rwlock_t *lock);
 #else /* SPLPAR */
 #define __spin_yield(x)	barrier()
 #define __rw_yield(x)	barrier()
-#define SHARED_PROCESSOR	0
 #endif
 
+static inline bool is_shared_processor(void)
+{
+/*
+ * LPPACA is only available on BOOK3S so guard anything LPPACA related to
+ * allow other platforms (which include this common header) to compile.
+ */
+#ifdef CONFIG_PPC_BOOK3S
+	return (IS_ENABLED(CONFIG_PPC_SPLPAR) &&
+		lppaca_shared_proc(local_paca->lppaca_ptr));
+#else
+	return false;
+#endif
+}
+
 static inline void arch_spin_lock(arch_spinlock_t *lock)
 {
 	while (1) {

@@ -117,7 +129,7 @@ static inline void arch_spin_lock(arch_spinlock_t *lock)
 			break;
 		do {
 			HMT_low();
-			if (SHARED_PROCESSOR)
+			if (is_shared_processor())
 				__spin_yield(lock);
 		} while (unlikely(lock->slock != 0));
 		HMT_medium();

@@ -136,7 +148,7 @@ void arch_spin_lock_flags(arch_spinlock_t *lock, unsigned long flags)
 		local_irq_restore(flags);
 		do {
 			HMT_low();
-			if (SHARED_PROCESSOR)
+			if (is_shared_processor())
 				__spin_yield(lock);
 		} while (unlikely(lock->slock != 0));
 		HMT_medium();

@@ -226,7 +238,7 @@ static inline void arch_read_lock(arch_rwlock_t *rw)
 			break;
 		do {
 			HMT_low();
-			if (SHARED_PROCESSOR)
+			if (is_shared_processor())
 				__rw_yield(rw);
 		} while (unlikely(rw->lock < 0));
 		HMT_medium();

@@ -240,7 +252,7 @@ static inline void arch_write_lock(arch_rwlock_t *rw)
 			break;
 		do {
 			HMT_low();
-			if (SHARED_PROCESSOR)
+			if (is_shared_processor())
 				__rw_yield(rw);
 		} while (unlikely(rw->lock != 0));
 		HMT_medium();

-- 
2.22.0

Booting w/ ppc64le_defconfig + CONFIG_PREEMPT results in the attached
kernel trace due to calling shared-processor spinlocks while not running
in an SPLPAR. Previously, the out-of-line spinlocks implementations were
selected based on CONFIG_PPC_SPLPAR at compile time without a runtime
shared-processor LPAR check.

To fix, call the actual spinlock implementations from a set of common
functions, spin_yield() and rw_yield(), which check for shared-processor
LPAR during runtime and select the appropriate lock implementation.

[    0.430878] BUG: Kernel NULL pointer dereference at 0x00000100
[    0.431991] Faulting instruction address: 0xc000000000097f88
[    0.432934] Oops: Kernel access of bad area, sig: 7 [#1]
[    0.433448] LE PAGE_SIZE=64K MMU=Radix MMU=Hash PREEMPT SMP NR_CPUS=2048 NUMA PowerNV
[    0.434479] Modules linked in:
[    0.435055] CPU: 0 PID: 2 Comm: kthreadd Not tainted 5.2.0-rc6-00491-g249155c20f9b #28
[    0.435730] NIP:  c000000000097f88 LR: c000000000c07a88 CTR: c00000000015ca10
[    0.436383] REGS: c0000000727079f0 TRAP: 0300   Not tainted  (5.2.0-rc6-00491-g249155c20f9b)
[    0.437004] MSR:  9000000002009033 <SF,HV,VEC,EE,ME,IR,DR,RI,LE>  CR: 84000424  XER: 20040000
[    0.437874] CFAR: c000000000c07a84 DAR: 0000000000000100 DSISR: 00080000 IRQMASK: 1
[    0.437874] GPR00: c000000000c07a88 c000000072707c80 c000000001546300 c00000007be38a80
[    0.437874] GPR04: c0000000726f0c00 0000000000000002 c00000007279c980 0000000000000100
[    0.437874] GPR08: c000000001581b78 0000000080000001 0000000000000008 c00000007279c9b0
[    0.437874] GPR12: 0000000000000000 c000000001730000 c000000000142558 0000000000000000
[    0.437874] GPR16: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
[    0.437874] GPR20: 0000000000000000 0000000000000000 0000000000000000 0000000000000000
[    0.437874] GPR24: c00000007be38a80 c000000000c002f4 0000000000000000 0000000000000000
[    0.437874] GPR28: c000000072221a00 c0000000726c2600 c00000007be38a80 c00000007be38a80
[    0.443992] NIP [c000000000097f88] __spin_yield+0x48/0xa0
[    0.444523] LR [c000000000c07a88] __raw_spin_lock+0xb8/0xc0
[    0.445080] Call Trace:
[    0.445670] [c000000072707c80] [c000000072221a00] 0xc000000072221a00 (unreliable)
[    0.446425] [c000000072707cb0] [c000000000bffb0c] __schedule+0xbc/0x850
[    0.447078] [c000000072707d70] [c000000000c002f4] schedule+0x54/0x130
[    0.447694] [c000000072707da0] [c0000000001427dc] kthreadd+0x28c/0x2b0
[    0.448389] [c000000072707e20] [c00000000000c1cc] ret_from_kernel_thread+0x5c/0x70
[    0.449143] Instruction dump:
[    0.449821] 4d9e0020 552a043e 210a07ff 79080fe0 0b080000 3d020004 3908b878 794a1f24
[    0.450587] e8e80000 7ce7502a e8e70000 38e70100 <7ca03c2c> 70a70001 78a50020 4d820020
[    0.452808] ---[ end trace 474d6b2b8fc5cb7e ]---

Signed-off-by: Christopher M. Riedl <redacted>
---
Changes since v2:
 - Directly call splpar_*_yield() to avoid duplicate call to
   is_shared_processor() in some cases

 arch/powerpc/include/asm/spinlock.h | 36 ++++++++++++++++++++---------
 1 file changed, 25 insertions(+), 11 deletions(-)

diff --git a/arch/powerpc/include/asm/spinlock.h b/arch/powerpc/include/asm/spinlock.h
index 0a8270183770..8935315c80ff 100644
--- a/arch/powerpc/include/asm/spinlock.h
+++ b/arch/powerpc/include/asm/spinlock.h

@@ -103,11 +103,9 @@ static inline int arch_spin_trylock(arch_spinlock_t *lock)
 /* We only yield to the hypervisor if we are in shared processor mode */
 void splpar_spin_yield(arch_spinlock_t *lock);
 void splpar_rw_yield(arch_rwlock_t *lock);
-#define __spin_yield(x) splpar_spin_yield(x)
-#define __rw_yield(x) splpar_rw_yield(x)
 #else /* SPLPAR */
-#define __spin_yield(x)	barrier()
-#define __rw_yield(x)	barrier()
+static inline void splpar_spin_yield(arch_spinlock_t *lock) {};
+static inline void splpar_rw_yield(arch_rwlock_t *lock) {};
 #endif
 
 static inline bool is_shared_processor(void)

@@ -124,6 +122,22 @@ static inline bool is_shared_processor(void)
 #endif
 }
 
+static inline void spin_yield(arch_spinlock_t *lock)
+{
+	if (is_shared_processor())
+		splpar_spin_yield(lock);
+	else
+		barrier();
+}
+
+static inline void rw_yield(arch_rwlock_t *lock)
+{
+	if (is_shared_processor())
+		splpar_rw_yield(lock);
+	else
+		barrier();
+}
+
 static inline void arch_spin_lock(arch_spinlock_t *lock)
 {
 	while (1) {

@@ -132,7 +146,7 @@ static inline void arch_spin_lock(arch_spinlock_t *lock)
 		do {
 			HMT_low();
 			if (is_shared_processor())
-				__spin_yield(lock);
+				splpar_spin_yield(lock);
 		} while (unlikely(lock->slock != 0));
 		HMT_medium();
 	}

@@ -151,7 +165,7 @@ void arch_spin_lock_flags(arch_spinlock_t *lock, unsigned long flags)
 		do {
 			HMT_low();
 			if (is_shared_processor())
-				__spin_yield(lock);
+				splpar_spin_yield(lock);
 		} while (unlikely(lock->slock != 0));
 		HMT_medium();
 		local_irq_restore(flags_dis);

@@ -241,7 +255,7 @@ static inline void arch_read_lock(arch_rwlock_t *rw)
 		do {
 			HMT_low();
 			if (is_shared_processor())
-				__rw_yield(rw);
+				splpar_rw_yield(rw);
 		} while (unlikely(rw->lock < 0));
 		HMT_medium();
 	}

@@ -255,7 +269,7 @@ static inline void arch_write_lock(arch_rwlock_t *rw)
 		do {
 			HMT_low();
 			if (is_shared_processor())
-				__rw_yield(rw);
+				splpar_rw_yield(rw);
 		} while (unlikely(rw->lock != 0));
 		HMT_medium();
 	}

@@ -295,9 +309,9 @@ static inline void arch_write_unlock(arch_rwlock_t *rw)
 	rw->lock = 0;
 }
 
-#define arch_spin_relax(lock)	__spin_yield(lock)
-#define arch_read_relax(lock)	__rw_yield(lock)
-#define arch_write_relax(lock)	__rw_yield(lock)
+#define arch_spin_relax(lock)	spin_yield(lock)
+#define arch_read_relax(lock)	rw_yield(lock)
+#define arch_write_relax(lock)	rw_yield(lock)
 
 /* See include/linux/spinlock.h */
 #define smp_mb__after_spinlock()   smp_mb()

-- 
2.22.0

The __rw_yield and __spin_yield locks only pertain to SPLPAR mode.
Rename them to make this relationship obvious.

Signed-off-by: Christopher M. Riedl <redacted>
Reviewed-by: Andrew Donnellan <redacted>
---
 arch/powerpc/include/asm/spinlock.h | 6 ++++--
 arch/powerpc/lib/locks.c            | 6 +++---
 2 files changed, 7 insertions(+), 5 deletions(-)

diff --git a/arch/powerpc/include/asm/spinlock.h b/arch/powerpc/include/asm/spinlock.h
index dc5fcea1f006..0a8270183770 100644
--- a/arch/powerpc/include/asm/spinlock.h
+++ b/arch/powerpc/include/asm/spinlock.h

@@ -101,8 +101,10 @@ static inline int arch_spin_trylock(arch_spinlock_t *lock)
 
 #if defined(CONFIG_PPC_SPLPAR)
 /* We only yield to the hypervisor if we are in shared processor mode */
-extern void __spin_yield(arch_spinlock_t *lock);
-extern void __rw_yield(arch_rwlock_t *lock);
+void splpar_spin_yield(arch_spinlock_t *lock);
+void splpar_rw_yield(arch_rwlock_t *lock);
+#define __spin_yield(x) splpar_spin_yield(x)
+#define __rw_yield(x) splpar_rw_yield(x)
 #else /* SPLPAR */
 #define __spin_yield(x)	barrier()
 #define __rw_yield(x)	barrier()

diff --git a/arch/powerpc/lib/locks.c b/arch/powerpc/lib/locks.c
index 6550b9e5ce5f..6440d5943c00 100644
--- a/arch/powerpc/lib/locks.c
+++ b/arch/powerpc/lib/locks.c

@@ -18,7 +18,7 @@
 #include <asm/hvcall.h>
 #include <asm/smp.h>
 
-void __spin_yield(arch_spinlock_t *lock)
+void splpar_spin_yield(arch_spinlock_t *lock)
 {
 	unsigned int lock_value, holder_cpu, yield_count;
 

@@ -36,14 +36,14 @@ void __spin_yield(arch_spinlock_t *lock)
 	plpar_hcall_norets(H_CONFER,
 		get_hard_smp_processor_id(holder_cpu), yield_count);
 }
-EXPORT_SYMBOL_GPL(__spin_yield);
+EXPORT_SYMBOL_GPL(splpar_spin_yield);
 
 /*
  * Waiting for a read lock or a write lock on a rwlock...
  * This turns out to be the same for read and write locks, since
  * we only know the holder if it is write-locked.
  */
-void __rw_yield(arch_rwlock_t *rw)
+void splpar_rw_yield(arch_rwlock_t *rw)
 {
 	int lock_value;
 	unsigned int holder_cpu, yield_count;

-- 
2.22.0

Hi "Christopher,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on linus/master]
[cannot apply to v5.3-rc4 next-20190809]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Christopher-M-Riedl/Fix-oops-in-shared-processor-spinlocks/20190806-204502
config: powerpc-powernv_defconfig (attached as .config)
compiler: powerpc64le-linux-gcc (GCC) 7.4.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        GCC_VERSION=7.4.0 make.cross ARCH=powerpc 

If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <redacted>

All errors (new ones prefixed by >>):

   In file included from include/linux/spinlock.h:89:0,
                    from include/linux/seqlock.h:36,
                    from include/linux/time.h:6,
                    from include/linux/compat.h:10,
                    from arch/powerpc/kernel/asm-offsets.c:14:
   arch/powerpc/include/asm/spinlock.h: In function 'is_shared_processor':

quoted

arch/powerpc/include/asm/spinlock.h:119:34: error: 'struct paca_struct' has no member named 'lppaca_ptr'; did you mean 'slb_cache_ptr'?

      lppaca_shared_proc(local_paca->lppaca_ptr));
                                     ^~~~~~~~~~
                                     slb_cache_ptr
   make[2]: *** [arch/powerpc/kernel/asm-offsets.s] Error 1
   make[2]: Target '__build' not remade because of errors.
   make[1]: *** [prepare0] Error 2
   make[1]: Target 'prepare' not remade because of errors.
   make: *** [sub-make] Error 2
   7 real  4 user  3 sys  110.24% cpu 	make prepare

vim +119 arch/powerpc/include/asm/spinlock.h

   110	
   111	static inline bool is_shared_processor(void)
   112	{
   113	/*
   114	 * LPPACA is only available on BOOK3S so guard anything LPPACA related to
   115	 * allow other platforms (which include this common header) to compile.
   116	 */
   117	#ifdef CONFIG_PPC_BOOK3S
   118		return (IS_ENABLED(CONFIG_PPC_SPLPAR) &&
 > 119			lppaca_shared_proc(local_paca->lppaca_ptr));
   120	#else
   121		return false;
   122	#endif
   123	}
   124	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

kbuild test robot [off-list ref] writes:

Hi "Christopher,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on linus/master]
[cannot apply to v5.3-rc4 next-20190809]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Christopher-M-Riedl/Fix-oops-in-shared-processor-spinlocks/20190806-204502
config: powerpc-powernv_defconfig (attached as .config)
compiler: powerpc64le-linux-gcc (GCC) 7.4.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        GCC_VERSION=7.4.0 make.cross ARCH=powerpc 

If you fix the issue, kindly add following tag
Reported-by: kbuild test robot <redacted>

All errors (new ones prefixed by >>):

   In file included from include/linux/spinlock.h:89:0,
                    from include/linux/seqlock.h:36,
                    from include/linux/time.h:6,
                    from include/linux/compat.h:10,
                    from arch/powerpc/kernel/asm-offsets.c:14:
   arch/powerpc/include/asm/spinlock.h: In function 'is_shared_processor':

quoted

quoted

arch/powerpc/include/asm/spinlock.h:119:34: error: 'struct paca_struct' has no member named 'lppaca_ptr'; did you mean 'slb_cache_ptr'?

      lppaca_shared_proc(local_paca->lppaca_ptr));
                                     ^~~~~~~~~~
                                     slb_cache_ptr
   make[2]: *** [arch/powerpc/kernel/asm-offsets.s] Error 1
   make[2]: Target '__build' not remade because of errors.
   make[1]: *** [prepare0] Error 2
   make[1]: Target 'prepare' not remade because of errors.
   make: *** [sub-make] Error 2
   7 real  4 user  3 sys  110.24% cpu 	make prepare

vim +119 arch/powerpc/include/asm/spinlock.h

   110	
   111	static inline bool is_shared_processor(void)
   112	{
   113	/*
   114	 * LPPACA is only available on BOOK3S so guard anything LPPACA related to
   115	 * allow other platforms (which include this common header) to compile.
   116	 */
   117	#ifdef CONFIG_PPC_BOOK3S

I think you should use PPC_PSERIES here and that will fix it.

cheers

   118		return (IS_ENABLED(CONFIG_PPC_SPLPAR) &&
 > 119			lppaca_shared_proc(local_paca->lppaca_ptr));
   120	#else
   121		return false;
   122	#endif
   123	}
   124	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation