Skip to contentSkip to conversation

fix error handling in load_module()

PATCH In review Kamalesh Babulal 4 participants 8 messages Sep 7, 2009 – Sep 22, 2009

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 fix error handling in load_module()

Hi Rusty,

During our testing following call trace was seen. The testcase was
to compile the kernel based on the distro config and try to insert all the modules compiled.

#!/bin/sh

for module in modprobe -l | tr '\n' ' '
do
insert_module=basename $module .ko
modprobe -v $insert_module
done

freq_table sputrace hvcserver axonram pmi ipv6 fuse ehea ib

Sep  7 15:46:04 mjs22lp5 kernel: mveth ibmvscsic scsi_transport_srp scsi_tgt
Sep  7 15:46:04 mjs22lp5 kernel: NIP: c0000000000ebba0 LR: c0000000000ee79c CTR: 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: REGS: c00000002c90b8e0 TRAP: 0700 Tainted: P      D     (2.6.31-rc8)
Sep  7 15:46:04 mjs22lp5 kernel: MSR: 8000000000029032 <EE,ME,CE,IR,DR> CR: 24222488  XER: 00000008
Sep  7 15:46:04 mjs22lp5 kernel: TASK = c00000002ff40000[9062] 'modprobe' THREAD: c00000002c908000 CPU: 0
Sep  7 15:46:04 mjs22lp5 kernel: GPR00: 0000000000000010 c00000002c90bb60 c000000001421e68 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR04: c000000000691a5c c00000000009f5c4 0000000000000000 c0000000167f6630
Sep  7 15:46:04 mjs22lp5 kernel: GPR08: c0000000167f72a4 000000000000031f c000000000bb9580 000000000000031e
Sep  7 15:46:04 mjs22lp5 kernel: GPR12: 800000000631b800 c0000000015a2600 0000000000000000 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR16: 0000000000000033 d00000000fb1f6d0 d00000000fb1fe50 000000000000000e
Sep  7 15:46:04 mjs22lp5 kernel: GPR20: d00000000fb1efb8 d00000000fb62260 d00000000fb00000 8000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR24: 0000000000000004 d00000000fb1f190 0000000000000035 fffffffffffffff4
Sep  7 15:46:04 mjs22lp5 kernel: GPR28: 0000000000000000 000000000000031e c00000000137def8 c00000002c90bb60
Sep  7 15:46:04 mjs22lp5 kernel: NIP [c0000000000ebba0] .percpu_modfree+0xe8/0x210
Sep  7 15:46:04 mjs22lp5 kernel: LR [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: Call Trace:
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bb60] [c00000002c90bc00] 0xc00000002c90bc00 (unreliable)
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bc00] [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bd90] [c0000000000ee988] .SyS_init_module+0x94/0x2ac
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90be30] [c0000000000084dc] syscall_exit+0x0/0x40
Sep  7 15:46:04 mjs22lp5 kernel: Instruction dump:
Sep  7 15:46:05 mjs22lp5 kernel: 48000038 e8080006 793d0020 39080004 78090020 2f800000 409c000c 7c0000d0
Sep  7 15:46:05 mjs22lp5 kernel: 78090020 7d4a4a14 393d0001 4200ffb0 <0fe00000> 48000000 38a30001 7f83e378
Sep  7 15:46:05 mjs22lp5 kernel: ---[ end trace 3c8bbdf1034c7f0d ]---

Once the percpu_modalloc fails, percpu_modfree(mod->refptr) is called on a NULL pointer. We try calling it on a NULL pointer. The following patch fixes the problem by introducing a check for mod->refptr before calling percpu_modfree.

--

 kernel/module.c |    3 ++-
 1 files changed, 2 insertions(+), 1 deletions(-)
kernel/module.c +2 -1
index 2d53718..7f89258 100644
@@ -2379,7 +2379,8 @@ static noinline struct module *load_module(void __user *umod,
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
  free_init:
-	percpu_modfree(mod->refptr);
+	if (mod->refptr)
+		percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:

Kamalesh

maintainer
#1 lore raw
Quoted 40 lines · On Mon, 7 Sep 2009 19:45:58 +0530 Kamalesh Babulal…

Hi Rusty,

During our testing following call trace was seen. The testcase was
to compile the kernel based on the distro config and try to insert all the modules compiled.

#!/bin/sh

for module in modprobe -l | tr '\n' ' '
do
insert_module=basename $module .ko
modprobe -v $insert_module
done

freq_table sputrace hvcserver axonram pmi ipv6 fuse ehea ib

Sep  7 15:46:04 mjs22lp5 kernel: mveth ibmvscsic scsi_transport_srp scsi_tgt
Sep  7 15:46:04 mjs22lp5 kernel: NIP: c0000000000ebba0 LR: c0000000000ee79c CTR: 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: REGS: c00000002c90b8e0 TRAP: 0700 Tainted: P      D     (2.6.31-rc8)
Sep  7 15:46:04 mjs22lp5 kernel: MSR: 8000000000029032 <EE,ME,CE,IR,DR> CR: 24222488  XER: 00000008
Sep  7 15:46:04 mjs22lp5 kernel: TASK = c00000002ff40000[9062] 'modprobe' THREAD: c00000002c908000 CPU: 0
Sep  7 15:46:04 mjs22lp5 kernel: GPR00: 0000000000000010 c00000002c90bb60 c000000001421e68 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR04: c000000000691a5c c00000000009f5c4 0000000000000000 c0000000167f6630
Sep  7 15:46:04 mjs22lp5 kernel: GPR08: c0000000167f72a4 000000000000031f c000000000bb9580 000000000000031e
Sep  7 15:46:04 mjs22lp5 kernel: GPR12: 800000000631b800 c0000000015a2600 0000000000000000 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR16: 0000000000000033 d00000000fb1f6d0 d00000000fb1fe50 000000000000000e
Sep  7 15:46:04 mjs22lp5 kernel: GPR20: d00000000fb1efb8 d00000000fb62260 d00000000fb00000 8000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR24: 0000000000000004 d00000000fb1f190 0000000000000035 fffffffffffffff4
Sep  7 15:46:04 mjs22lp5 kernel: GPR28: 0000000000000000 000000000000031e c00000000137def8 c00000002c90bb60
Sep  7 15:46:04 mjs22lp5 kernel: NIP [c0000000000ebba0] .percpu_modfree+0xe8/0x210
Sep  7 15:46:04 mjs22lp5 kernel: LR [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: Call Trace:
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bb60] [c00000002c90bc00] 0xc00000002c90bc00 (unreliable)
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bc00] [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bd90] [c0000000000ee988] .SyS_init_module+0x94/0x2ac
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90be30] [c0000000000084dc] syscall_exit+0x0/0x40
Sep  7 15:46:04 mjs22lp5 kernel: Instruction dump:
Sep  7 15:46:05 mjs22lp5 kernel: 48000038 e8080006 793d0020 39080004 78090020 2f800000 409c000c 7c0000d0
Sep  7 15:46:05 mjs22lp5 kernel: 78090020 7d4a4a14 393d0001 4200ffb0 <0fe00000> 48000000 38a30001 7f83e378
Sep  7 15:46:05 mjs22lp5 kernel: ---[ end trace 3c8bbdf1034c7f0d ]---
Quoted 3 lines

Once the percpu_modalloc fails, percpu_modfree(mod->refptr) is called on a NULL pointer. We try calling it on a NULL pointer. The following patch fixes the problem by introducing a check for mod->refptr before calling percpu_modfree.

Where did it crash and why did it crash? That trace is pretty unclear.

Quoted 5 lines
kernel/module.c +0 -0
index 2d53718..7f89258 100644

kernel/module.c static noinline struct module *load_module(void __user *umod, · line 2386

Quoted 9 lines
kernel/module.c +2 -1
@@ -2379,7 +2379,8 @@ static noinline struct module *load_module(void __user *umod,
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
  free_init:
-	percpu_modfree(mod->refptr);
+	if (mod->refptr)
+		percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

It calls free_percpu(), all implementations of which appear to secretly support free_percpu(NULL).

So why did your machine crash?

This:

void free_percpu(void *ptr)

{
	void *addr = __pcpu_ptr_to_addr(ptr);
	struct pcpu_chunk *chunk;
	unsigned long flags;
	int off;

	if (!ptr)
		return;

is dangerous. The implementation of __pcpu_ptr_to_addr() can be overridden by asm/percpu.h and there's no reason why the compiler won't choose to pass a NULL into __pcpu_ptr_to_addr().

But there doesn't appear to be any overriding of __pcpu_ptr_to_addr() in 2.6.31 and the default __pcpu_ptr_to_addr() looks like it will handle a NULL pointer OK.

So again, why did your machine crash?

From: Andrew Morton <[email protected]>

__pcpu_ptr_to_addr() can be overridden by the architecture and might not behave well if passed a NULL pointer. So avoid calling it until we have verified that its arg is not NULL.

Cc Rusty Russell <[email protected]>
Cc Kamalesh Babulal <[email protected]>

---

 mm/percpu.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
mm/percpu.c renamed from mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull +3 -1
diff -puN mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull mm/percpu.c
@@ -957,7 +957,7 @@ static void pcpu_reclaim(struct work_str
  */
 void free_percpu(void *ptr)
 {
-	void *addr = __pcpu_ptr_to_addr(ptr);
+	void *addr;
 	struct pcpu_chunk *chunk;
 	unsigned long flags;
 	int off;
@@ -965,6 +965,8 @@ void free_percpu(void *ptr)
 	if (!ptr)
 		return;
 
+	addr = __pcpu_ptr_to_addr(ptr);
+
 	spin_lock_irqsave(&pcpu_lock, flags);
 
 	chunk = pcpu_chunk_addr_search(addr);

_

#2 lore raw
Quoted 17 lines · Andrew Morton
Quoted 14 lines · On Mon, 7 Sep 2009 19:45:58 +0530 Kamalesh Babulal…

Hi Rusty,

During our testing following call trace was seen. The testcase was
to compile the kernel based on the distro config and try to insert all the modules compiled.

#!/bin/sh

for module in modprobe -l | tr '\n' ' '
do
insert_module=basename $module .ko
modprobe -v $insert_module
done

<snip>

Quoted 7 lines
Quoted 4 lines

Once the percpu_modalloc fails, percpu_modfree(mod->refptr) is called on a NULL pointer. We try calling it on a NULL pointer. The following patch fixes the problem by introducing a check for mod->refptr before calling percpu_modfree.

Where did it crash and why did it crash? That trace is pretty unclear.

This was on the powerpc and used to crash after inserting around 790 modules, using the above script. My bad, sorry for unclear call trace. The kernel was configured with

CONFIG_SMP=y
CONFIG_MODULE_UNLOAD=y
CONFIG_HAVE_DYNAMIC_PER_CPU_AREA=n

Sep  7 15:46:04 mjs22lp5 kernel: yealink: yld-20051230:Yealink phone driver
Sep  7 15:46:04 mjs22lp5 kernel: Could not allocate 8 bytes percpu data
Sep  7 15:46:04 mjs22lp5 kernel: ------------[ cut here ]------------
Sep  7 15:46:04 mjs22lp5 kernel: kernel BUG at kernel/module.c:495!
Sep  7 15:46:04 mjs22lp5 kernel: Oops: Exception in kernel mode, sig: 5 [#8]
Sep  7 15:46:04 mjs22lp5 kernel: SMP NR_CPUS=2048 NUMA pSeries
Sep  7 15:46:04 mjs22lp5 kernel: Modules linked in: <snipped the modules name>
.
.
Sep  7 15:46:04 mjs22lp5 kernel: mveth ibmvscsic scsi_transport_srp scsi_tgt
Sep  7 15:46:04 mjs22lp5 kernel: NIP: c0000000000ebba0 LR: c0000000000ee79c CTR: 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: REGS: c00000002c90b8e0 TRAP: 0700 Tainted: P      D     (2.6.31-rc8)
Sep  7 15:46:04 mjs22lp5 kernel: MSR: 8000000000029032 <EE,ME,CE,IR,DR> CR: 24222488  XER: 00000008
Sep  7 15:46:04 mjs22lp5 kernel: TASK = c00000002ff40000[9062] 'modprobe' THREAD: c00000002c908000 CPU: 0
Sep  7 15:46:04 mjs22lp5 kernel: GPR00: 0000000000000010 c00000002c90bb60 c000000001421e68 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR04: c000000000691a5c c00000000009f5c4 0000000000000000 c0000000167f6630
Sep  7 15:46:04 mjs22lp5 kernel: GPR08: c0000000167f72a4 000000000000031f c000000000bb9580 000000000000031e
Sep  7 15:46:04 mjs22lp5 kernel: GPR12: 800000000631b800 c0000000015a2600 0000000000000000 0000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR16: 0000000000000033 d00000000fb1f6d0 d00000000fb1fe50 000000000000000e
Sep  7 15:46:04 mjs22lp5 kernel: GPR20: d00000000fb1efb8 d00000000fb62260 d00000000fb00000 8000000000000000
Sep  7 15:46:04 mjs22lp5 kernel: GPR24: 0000000000000004 d00000000fb1f190 0000000000000035 fffffffffffffff4
Sep  7 15:46:04 mjs22lp5 kernel: GPR28: 0000000000000000 000000000000031e c00000000137def8 c00000002c90bb60
Sep  7 15:46:04 mjs22lp5 kernel: NIP [c0000000000ebba0] .percpu_modfree+0xe8/0x210
Sep  7 15:46:04 mjs22lp5 kernel: LR [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: Call Trace:
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bb60] [c00000002c90bc00] 0xc00000002c90bc00 (unreliable)
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bc00] [c0000000000ee79c] .load_module+0x14f8/0x1650
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90bd90] [c0000000000ee988] .SyS_init_module+0x94/0x2ac
Sep  7 15:46:04 mjs22lp5 kernel: [c00000002c90be30] [c0000000000084dc] syscall_exit+0x0/0x40
Sep  7 15:46:04 mjs22lp5 kernel: Instruction dump:
Sep  7 15:46:05 mjs22lp5 kernel: 48000038 e8080006 793d0020 39080004 78090020 2f800000 409c000c 7c0000d0
Sep  7 15:46:05 mjs22lp5 kernel: 78090020 7d4a4a14 393d0001 4200ffb0 <0fe00000> 48000000 38a30001 7f83e378
Sep  7 15:46:05 mjs22lp5 kernel: ---[ end trace 3c8bbdf1034c7f0d ]---
Quoted 40 lines
Quoted 5 lines
kernel/module.c +0 -0
index 2d53718..7f89258 100644

kernel/module.c static noinline struct module *load_module(void __user *umod, · line 2386

Quoted 9 lines
kernel/module.c +2 -1
@@ -2379,7 +2379,8 @@ static noinline struct module *load_module(void __user *umod,
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
  free_init:
-	percpu_modfree(mod->refptr);
+	if (mod->refptr)
+		percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

It calls free_percpu(), all implementations of which appear to secretly support free_percpu(NULL).

So why did your machine crash?

This:

void free_percpu(void *ptr)

{
	void *addr = __pcpu_ptr_to_addr(ptr);
	struct pcpu_chunk *chunk;
	unsigned long flags;
	int off;

	if (!ptr)
		return;

is dangerous. The implementation of __pcpu_ptr_to_addr() can be overridden by asm/percpu.h and there's no reason why the compiler won't choose to pass a NULL into __pcpu_ptr_to_addr().

Quoted 6 lines

But there doesn't appear to be any overriding of __pcpu_ptr_to_addr() in 2.6.31 and the default __pcpu_ptr_to_addr() looks like it will handle a NULL pointer OK.

So again, why did your machine crash?

with CONFIG_HAVE_DYNAMIC_PER_CPU_AREA disabled, the below code path is called and the BUG() is hit,

static void percpu_modfree(void *freeme)

{
		unsigned int i;
		void *ptr = __per_cpu_start + block_size(pcpu_size[0]);
		int cpu; 

		/* First entry is core kernel percpu data. */
		for (i = 1; i < pcpu_num_used; ptr +=
						block_size(pcpu_size[i]), i++) {
				if (ptr == freeme) {
						pcpu_size[i] = -pcpu_size[i];
						goto free;
				}    
		}    
		BUG(); 

In the patch, the check was introduced before calling percpu_modfree.

Quoted 37 lines

From: Andrew Morton <[email protected]>

__pcpu_ptr_to_addr() can be overridden by the architecture and might not behave well if passed a NULL pointer. So avoid calling it until we have verified that its arg is not NULL.

Cc Rusty Russell <[email protected]>
Cc Kamalesh Babulal <[email protected]>

---

 mm/percpu.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff -puN mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull mm/percpu.c

mm/percpu.c renamed from mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull +3 -1
@@ -957,7 +957,7 @@ static void pcpu_reclaim(struct work_str
  */
 void free_percpu(void *ptr)
 {
-	void *addr = __pcpu_ptr_to_addr(ptr);
+	void *addr;
 	struct pcpu_chunk *chunk;
 	unsigned long flags;
 	int off;
@@ -965,6 +965,8 @@ void free_percpu(void *ptr)
 	if (!ptr)
 		return;
 
+	addr = __pcpu_ptr_to_addr(ptr);
+
 	spin_lock_irqsave(&pcpu_lock, flags);
 
 	chunk = pcpu_chunk_addr_search(addr);
Quoted 2 lines

_

Kamalesh

#3 lore raw
In reply to Andrew Morton
Quoted 18 lines · Andrew Morton
Quoted 3 lines · On Mon, 7 Sep 2009 19:45:58 +0530 Kamalesh Babulal…

Once the percpu_modalloc fails, percpu_modfree(mod->refptr) is called on a NULL pointer. We try calling it on a NULL pointer. The following patch fixes the problem by introducing a check for mod->refptr before calling percpu_modfree.

Where did it crash and why did it crash? That trace is pretty unclear.

Quoted 10 lines
kernel/module.c +1 -1
index 2d53718..7f89258 100644
@@ -2379,7 +2379,8 @@ static noinline struct module *load_module(void __user *umod,
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
  free_init:
-	percpu_modfree(mod->refptr);
+	if (mod->refptr)
Quoted 8 lines

kernel/module.c static noinline struct module *load_module(void __user *umod, · line 2386

Quoted 4 lines
kernel/module.c +1 -0
+		percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

Yes, percpu_modfree() should handle NULL. If it doesn't, that's the bug.

Confused,
Rusty.

maintainer
#4 lore raw

Hello, Rusty.

Quoted 5 lines · Rusty Russell
Quoted 3 lines

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

Yes, percpu_modfree() should handle NULL. If it doesn't, that's the bug.

Hmm... It seems the report was against pre-2.6.32-rcX window. ppc is still using the original percpu_modfree() and I don't think the it ever supported NULL free. In the current linus tree, the triggering BUG() would be at line 500 which is inside percpu_modfree() which checks @freeme matches any allocated address and if not triggers BUG().

I think something like the following should fix it.

 kernel/module.c |    3 +++
 1 file changed, 3 insertions(+)
kernel/module.c +3 -0
index 05ce49c..eed1f9e 100644
@@ -490,6 +490,9 @@ static void percpu_modfree(void *freeme)
 	void *ptr = __per_cpu_start + block_size(pcpu_size[0]);
 	int cpu;
 
+	if (unlikely(!freeme))
+		return;
+
 	/* First entry is core kernel percpu data. */
 	for (i = 1; i < pcpu_num_used; ptr += block_size(pcpu_size[i]), i++) {
 		if (ptr == freeme) {
Signature · 1 line
tejun
maintainer
#5 lore raw
In reply to Andrew Morton

Hello, Andrew.

Quoted 6 lines · Andrew Morton

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

It calls free_percpu(), all implementations of which appear to secretly support free_percpu(NULL).

Eh... unfortunately, the original percpu_modfree() implementation didn't seem to support it.

Quoted 51 lines

So why did your machine crash?

This:

void free_percpu(void *ptr)

{
	void *addr = __pcpu_ptr_to_addr(ptr);
	struct pcpu_chunk *chunk;
	unsigned long flags;
	int off;

	if (!ptr)
		return;

is dangerous. The implementation of __pcpu_ptr_to_addr() can be overridden by asm/percpu.h and there's no reason why the compiler won't choose to pass a NULL into __pcpu_ptr_to_addr().

But there doesn't appear to be any overriding of __pcpu_ptr_to_addr() in 2.6.31 and the default __pcpu_ptr_to_addr() looks like it will handle a NULL pointer OK.

So again, why did your machine crash?

From: Andrew Morton <[email protected]>

__pcpu_ptr_to_addr() can be overridden by the architecture and might not behave well if passed a NULL pointer. So avoid calling it until we have verified that its arg is not NULL.

Cc Rusty Russell <[email protected]>
Cc Kamalesh Babulal <[email protected]>

---

 mm/percpu.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff -puN mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull mm/percpu.c

mm/percpu.c renamed from mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull +1 -1
@@ -957,7 +957,7 @@ static void pcpu_reclaim(struct work_str
  */
 void free_percpu(void *ptr)
 {
-	void *addr = __pcpu_ptr_to_addr(ptr);
+	void *addr;
 	struct pcpu_chunk *chunk;
 	unsigned long flags;
 	int off;

mm/percpu.c void free_percpu(void *ptr) · line 969

Quoted 5 lines
mm/percpu.c renamed from mm/percpu.c~percpu-avoid-calling-__pcpu_ptr_to_addrnull +2 -0
@@ -965,6 +965,8 @@ void free_percpu(void *ptr)
 	if (!ptr)
 		return;
 
+	addr = __pcpu_ptr_to_addr(ptr);
+

__pcpu_ptr_to_addr() and reverse should be simple arithmetic transformations. The sole reason why it's defined as overridable was mostly because I didn't know whether all archs could be unified to use the same macro (and different variants were used early during development) but yeap no harm in being careful.

Thanks.

Signature · 1 line
tejun
#6 lore raw
Quoted 7 lines · Tejun Heo

Hello, Andrew.

Quoted 4 lines · Andrew Morton

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

Quoted 5 lines
Quoted 2 lines

It calls free_percpu(), all implementations of which appear to secretly support free_percpu(NULL).

Eh... unfortunately, the original percpu_modfree() implementation didn't seem to support it.

OK, I'll Andrew's fix for Tejun, and after his (spot-on!) comment about percpu_modfree never taking NULL, I've fixed the one caller to match the other two:

Subject: module: don't call percpu_modfree on NULL pointer.

The general one handles NULL, the static obsolescent (CONFIG_HAVE_LEGACY_PER_CPU_AREA) one in module.c doesn't; Eric's commit 720eba31 assumed it did, and various frobbings since then kept that assumption.

All other callers in module.c all protect it with an if; this effectively does the same as free_init is only goto if we fail percpu_modalloc().

Cc Eric Dumazet <[email protected]>
Cc Masami Hiramatsu <[email protected]>
Cc Américo Wang <[email protected]>
kernel/module.c +1 -1
@@ -2522,8 +2522,8 @@ static noinline struct module *load_modu
  free_unload:
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
+	percpu_modfree(mod->refptr);
  free_init:
-	percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:
#7 lore raw
Quoted 20 lines · Rusty Russell
Quoted 7 lines · Tejun Heo

Hello, Andrew.

Quoted 4 lines · Andrew Morton

My reverse engineering of the secret, undocumented percpu_modfree() indicates that its mad inventor intended that percpu_modfree(NULL) be a valid thing to do.

Quoted 5 lines
Quoted 2 lines

It calls free_percpu(), all implementations of which appear to secretly support free_percpu(NULL).

Eh... unfortunately, the original percpu_modfree() implementation didn't seem to support it.

OK, I'll Andrew's fix for Tejun, and after his (spot-on!) comment about percpu_modfree never taking NULL, I've fixed the one caller to match the other two:

Subject: module: don't call percpu_modfree on NULL pointer.

Quoted 7 lines

The general one handles NULL, the static obsolescent (CONFIG_HAVE_LEGACY_PER_CPU_AREA) one in module.c doesn't; Eric's commit 720eba31 assumed it did, and various frobbings since then kept that assumption.

All other callers in module.c all protect it with an if; this effectively does the same as free_init is only goto if we fail percpu_modalloc().

Thanks, the patch fixes the issue.

Quoted 20 lines
Cc Eric Dumazet <[email protected]>
Cc Masami Hiramatsu <[email protected]>
Cc Américo Wang <[email protected]>
kernel/module.c +1 -2
@@ -2522,8 +2522,8 @@ static noinline struct module *load_modu
  free_unload:
 	module_unload_free(mod);
 #if defined(CONFIG_MODULE_UNLOAD) && defined(CONFIG_SMP)
+	percpu_modfree(mod->refptr);
  free_init:
-	percpu_modfree(mod->refptr);
 #endif
 	module_free(mod, mod->module_init);
  free_core:
--
Quoted 4 lines

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

Kamalesh