Skip to contentSkip to conversation

libnvdimm/region: Initialize bad block for volatile namespaces

PATCH Superseded Aneesh Kumar K.V 2 participants 4 messages Sep 17, 2019 – Sep 19, 2019 libnvdimm: non-volatile memory device

v2 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 libnvdimm/region: Initialize bad block for volatile namespaces

We do check for a bad block during namespace init and that use region bad block list. We need to initialize the bad block for volatile regions for this to work. We also observe a lockdep warning as below because the lock is not initialized correctly since we skip bad block init for volatile regions.

INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.

 CPU: 2 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc1-15699-g3dee241c937e #149
 Call Trace:
 [c0000000f95cb250] [c00000000147dd84] dump_stack+0xe8/0x164 (unreliable)
 [c0000000f95cb2a0] [c00000000022ccd8] register_lock_class+0x308/0xa60
 [c0000000f95cb3a0] [c000000000229cc0] __lock_acquire+0x170/0x1ff0
 [c0000000f95cb4c0] [c00000000022c740] lock_acquire+0x220/0x270
 [c0000000f95cb580] [c000000000a93230] badblocks_check+0xc0/0x290
 [c0000000f95cb5f0] [c000000000d97540] nd_pfn_validate+0x5c0/0x7f0
 [c0000000f95cb6d0] [c000000000d98300] nd_dax_probe+0xd0/0x1f0
 [c0000000f95cb760] [c000000000d9b66c] nd_pmem_probe+0x10c/0x160
 [c0000000f95cb790] [c000000000d7f5ec] nvdimm_bus_probe+0x10c/0x240
 [c0000000f95cb820] [c000000000d0f844] really_probe+0x254/0x4e0
 [c0000000f95cb8b0] [c000000000d0fdfc] driver_probe_device+0x16c/0x1e0
 [c0000000f95cb930] [c000000000d10238] device_driver_attach+0x68/0xa0
 [c0000000f95cb970] [c000000000d1040c] __driver_attach+0x19c/0x1c0
 [c0000000f95cb9f0] [c000000000d0c4c4] bus_for_each_dev+0x94/0x130
 [c0000000f95cba50] [c000000000d0f014] driver_attach+0x34/0x50
 [c0000000f95cba70] [c000000000d0e208] bus_add_driver+0x178/0x2f0
 [c0000000f95cbb00] [c000000000d117c8] driver_register+0x108/0x170
 [c0000000f95cbb70] [c000000000d7edb0] __nd_driver_register+0xe0/0x100
 [c0000000f95cbbd0] [c000000001a6baa4] nd_pmem_driver_init+0x34/0x48
 [c0000000f95cbbf0] [c0000000000106f4] do_one_initcall+0x1d4/0x4b0
 [c0000000f95cbcd0] [c0000000019f499c] kernel_init_freeable+0x544/0x65c
 [c0000000f95cbdb0] [c000000000010d6c] kernel_init+0x2c/0x180
 [c0000000f95cbe20] [c00000000000b954] ret_from_kernel_thread+0x5c/0x68

---
Changes from V1:
* update commit subject

 drivers/nvdimm/region.c      | 4 ++--
 drivers/nvdimm/region_devs.c | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/nvdimm/region.c +2 -2
index 37bf8719a2a4..0f6978e72e7c 100644
@@ -34,7 +34,7 @@ static int nd_region_probe(struct device *dev)
 	if (rc)
 		return rc;
 
-	if (is_nd_pmem(&nd_region->dev)) {
+	if (is_memory(&nd_region->dev)) {
 		struct resource ndr_res;
 
 		if (devm_init_badblocks(dev, &nd_region->bb))
@@ -123,7 +123,7 @@ static void nd_region_notify(struct device *dev, enum nvdimm_event event)
 		struct nd_region *nd_region = to_nd_region(dev);
 		struct resource res;
 
-		if (is_nd_pmem(&nd_region->dev)) {
+		if (is_memory(&nd_region->dev)) {
 			res.start = nd_region->ndr_start;
 			res.end = nd_region->ndr_start +
 				nd_region->ndr_size - 1;
drivers/nvdimm/region_devs.c +1 -1
index 9550202aa7be..f462b1354ab5 100644
@@ -632,7 +632,7 @@ static umode_t region_visible(struct kobject *kobj, struct attribute *a, int n)
 	if (!is_memory(dev) && a == &dev_attr_dax_seed.attr)
 		return 0;
 
-	if (!is_nd_pmem(dev) && a == &dev_attr_badblocks.attr)
+	if (!is_memory(dev) && a == &dev_attr_badblocks.attr)
 		return 0;
 
 	if (a == &dev_attr_resource.attr) {
Signature · 6 lines
2.21.0

_______________________________________________
Linux-nvdimm mailing list
[email protected]
https://lists.01.org/mailman/listinfo/linux-nvdimm
#1 lore raw
Quoted 36 lines · Aneesh Kumar K.V

We do check for a bad block during namespace init and that use region bad block list. We need to initialize the bad block for volatile regions for this to work. We also observe a lockdep warning as below because the lock is not initialized correctly since we skip bad block init for volatile regions.

INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.

 CPU: 2 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc1-15699-g3dee241c937e #149
 Call Trace:
 [c0000000f95cb250] [c00000000147dd84] dump_stack+0xe8/0x164 (unreliable)
 [c0000000f95cb2a0] [c00000000022ccd8] register_lock_class+0x308/0xa60
 [c0000000f95cb3a0] [c000000000229cc0] __lock_acquire+0x170/0x1ff0
 [c0000000f95cb4c0] [c00000000022c740] lock_acquire+0x220/0x270
 [c0000000f95cb580] [c000000000a93230] badblocks_check+0xc0/0x290
 [c0000000f95cb5f0] [c000000000d97540] nd_pfn_validate+0x5c0/0x7f0
 [c0000000f95cb6d0] [c000000000d98300] nd_dax_probe+0xd0/0x1f0
 [c0000000f95cb760] [c000000000d9b66c] nd_pmem_probe+0x10c/0x160
 [c0000000f95cb790] [c000000000d7f5ec] nvdimm_bus_probe+0x10c/0x240
 [c0000000f95cb820] [c000000000d0f844] really_probe+0x254/0x4e0
 [c0000000f95cb8b0] [c000000000d0fdfc] driver_probe_device+0x16c/0x1e0
 [c0000000f95cb930] [c000000000d10238] device_driver_attach+0x68/0xa0
 [c0000000f95cb970] [c000000000d1040c] __driver_attach+0x19c/0x1c0
 [c0000000f95cb9f0] [c000000000d0c4c4] bus_for_each_dev+0x94/0x130
 [c0000000f95cba50] [c000000000d0f014] driver_attach+0x34/0x50
 [c0000000f95cba70] [c000000000d0e208] bus_add_driver+0x178/0x2f0
 [c0000000f95cbb00] [c000000000d117c8] driver_register+0x108/0x170
 [c0000000f95cbb70] [c000000000d7edb0] __nd_driver_register+0xe0/0x100
 [c0000000f95cbbd0] [c000000001a6baa4] nd_pmem_driver_init+0x34/0x48
 [c0000000f95cbbf0] [c0000000000106f4] do_one_initcall+0x1d4/0x4b0
 [c0000000f95cbcd0] [c0000000019f499c] kernel_init_freeable+0x544/0x65c
 [c0000000f95cbdb0] [c000000000010d6c] kernel_init+0x2c/0x180
 [c0000000f95cbe20] [c00000000000b954] ret_from_kernel_thread+0x5c/0x68
Quoted 4 lines

---
Changes from V1:
* update commit subject

What about the is_nd_pmem() call in nvdimm_clear_badblocks_region()?
_______________________________________________
Linux-nvdimm mailing list
[email protected]
https://lists.01.org/mailman/listinfo/linux-nvdimm

#2 lore raw
Quoted 38 lines · Dan Williams
Quoted 36 lines · Aneesh Kumar K.V

We do check for a bad block during namespace init and that use region bad block list. We need to initialize the bad block for volatile regions for this to work. We also observe a lockdep warning as below because the lock is not initialized correctly since we skip bad block init for volatile regions.

INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.

  CPU: 2 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc1-15699-g3dee241c937e #149
  Call Trace:
  [c0000000f95cb250] [c00000000147dd84] dump_stack+0xe8/0x164 (unreliable)
  [c0000000f95cb2a0] [c00000000022ccd8] register_lock_class+0x308/0xa60
  [c0000000f95cb3a0] [c000000000229cc0] __lock_acquire+0x170/0x1ff0
  [c0000000f95cb4c0] [c00000000022c740] lock_acquire+0x220/0x270
  [c0000000f95cb580] [c000000000a93230] badblocks_check+0xc0/0x290
  [c0000000f95cb5f0] [c000000000d97540] nd_pfn_validate+0x5c0/0x7f0
  [c0000000f95cb6d0] [c000000000d98300] nd_dax_probe+0xd0/0x1f0
  [c0000000f95cb760] [c000000000d9b66c] nd_pmem_probe+0x10c/0x160
  [c0000000f95cb790] [c000000000d7f5ec] nvdimm_bus_probe+0x10c/0x240
  [c0000000f95cb820] [c000000000d0f844] really_probe+0x254/0x4e0
  [c0000000f95cb8b0] [c000000000d0fdfc] driver_probe_device+0x16c/0x1e0
  [c0000000f95cb930] [c000000000d10238] device_driver_attach+0x68/0xa0
  [c0000000f95cb970] [c000000000d1040c] __driver_attach+0x19c/0x1c0
  [c0000000f95cb9f0] [c000000000d0c4c4] bus_for_each_dev+0x94/0x130
  [c0000000f95cba50] [c000000000d0f014] driver_attach+0x34/0x50
  [c0000000f95cba70] [c000000000d0e208] bus_add_driver+0x178/0x2f0
  [c0000000f95cbb00] [c000000000d117c8] driver_register+0x108/0x170
  [c0000000f95cbb70] [c000000000d7edb0] __nd_driver_register+0xe0/0x100
  [c0000000f95cbbd0] [c000000001a6baa4] nd_pmem_driver_init+0x34/0x48
  [c0000000f95cbbf0] [c0000000000106f4] do_one_initcall+0x1d4/0x4b0
  [c0000000f95cbcd0] [c0000000019f499c] kernel_init_freeable+0x544/0x65c
  [c0000000f95cbdb0] [c000000000010d6c] kernel_init+0x2c/0x180
  [c0000000f95cbe20] [c00000000000b954] ret_from_kernel_thread+0x5c/0x68
Quoted 7 lines
Quoted 4 lines

---
Changes from V1:
* update commit subject

What about the is_nd_pmem() call in nvdimm_clear_badblocks_region()?

Missed that. Yes that also needs an updatet. Will you be able to update that or you want me to send a V3?

-aneesh
_______________________________________________
Linux-nvdimm mailing list
[email protected]
https://lists.01.org/mailman/listinfo/linux-nvdimm

#3 lore raw
Quoted 46 lines · Aneesh Kumar K.V
Quoted 9 lines · Dan Williams
Quoted 7 lines · Aneesh Kumar K.V

We do check for a bad block during namespace init and that use region bad block list. We need to initialize the bad block for volatile regions for this to work. We also observe a lockdep warning as below because the lock is not initialized correctly since we skip bad block init for volatile regions.

Quoted 4 lines
Quoted 4 lines

INFO: trying to register non-static key.
the code is fine but needs lockdep annotation.
turning off the locking correctness validator.

  CPU: 2 PID: 1 Comm: swapper/0 Not tainted

5.3.0-rc1-15699-g3dee241c937e #149

Quoted 2 lines
Quoted 2 lines
  Call Trace:
  [c0000000f95cb250] [c00000000147dd84] dump_stack+0xe8/0x164

(unreliable)

Quoted 28 lines
Quoted 27 lines

[c0000000f95cb2a0] [c00000000022ccd8] register_lock_class+0x308/0xa60 [c0000000f95cb3a0] [c000000000229cc0] __lock_acquire+0x170/0x1ff0 [c0000000f95cb4c0] [c00000000022c740] lock_acquire+0x220/0x270 [c0000000f95cb580] [c000000000a93230] badblocks_check+0xc0/0x290 [c0000000f95cb5f0] [c000000000d97540] nd_pfn_validate+0x5c0/0x7f0 [c0000000f95cb6d0] [c000000000d98300] nd_dax_probe+0xd0/0x1f0 [c0000000f95cb760] [c000000000d9b66c] nd_pmem_probe+0x10c/0x160 [c0000000f95cb790] [c000000000d7f5ec] nvdimm_bus_probe+0x10c/0x240 [c0000000f95cb820] [c000000000d0f844] really_probe+0x254/0x4e0 [c0000000f95cb8b0] [c000000000d0fdfc] driver_probe_device+0x16c/0x1e0 [c0000000f95cb930] [c000000000d10238] device_driver_attach+0x68/0xa0 [c0000000f95cb970] [c000000000d1040c] __driver_attach+0x19c/0x1c0 [c0000000f95cb9f0] [c000000000d0c4c4] bus_for_each_dev+0x94/0x130 [c0000000f95cba50] [c000000000d0f014] driver_attach+0x34/0x50 [c0000000f95cba70] [c000000000d0e208] bus_add_driver+0x178/0x2f0 [c0000000f95cbb00] [c000000000d117c8] driver_register+0x108/0x170 [c0000000f95cbb70] [c000000000d7edb0] __nd_driver_register+0xe0/0x100 [c0000000f95cbbd0] [c000000001a6baa4] nd_pmem_driver_init+0x34/0x48 [c0000000f95cbbf0] [c0000000000106f4] do_one_initcall+0x1d4/0x4b0 [c0000000f95cbcd0] [c0000000019f499c] kernel_init_freeable+0x544/0x65c [c0000000f95cbdb0] [c000000000010d6c] kernel_init+0x2c/0x180 [c0000000f95cbe20] [c00000000000b954] ret_from_kernel_thread+0x5c/0x68

---
Changes from V1:
* update commit subject

Quoted 6 lines
Quoted 2 lines

What about the is_nd_pmem() call in nvdimm_clear_badblocks_region()?

Missed that. Yes that also needs an updatet. Will you be able to update that or you want me to send a V3?

v3 please so it will get superseded in patchwork, and can keep the correlation between patchwork id and kernel commit.

(sent from my phone, forgive the HTML mail)
_______________________________________________
Linux-nvdimm mailing list
[email protected]
https://lists.01.org/mailman/listinfo/linux-nvdimm