[PATCH] [v2] bcache: stop using the deprecated get_seconds()

[PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Arnd Bergmann]

The get_seconds function is deprecated now since it returns a 32-bit
value that will eventually overflow, and we are replacing it throughout
the kernel with ktime_get_seconds() or ktime_get_real_seconds() that
return a time64_t.

bcache uses get_seconds() to read the current system time and store it in
the superblock as well as in uuid_entry structures that are user visible.

Unfortunately, the two structures in are still limited to 32 bits, so this
won't fix any real problems but will still overflow in year 2106. Let's
at least document that properly, in case we get an updated format in the
future it can be fixed. We still have a long time before the overflow
and checking the tools at https://github.com/koverstreet/bcache-tools
reveals no access to any of them.

Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 drivers/md/bcache/super.c   | 12 ++++++------
 include/uapi/linux/bcache.h |  4 ++--
 2 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
index fa4058e43202..74746d8ee05e 100644
--- a/drivers/md/bcache/super.c
+++ b/drivers/md/bcache/super.c
@@ -181,7 +181,7 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
 		goto err;
 	}
 
-	sb->last_mount = get_seconds();
+	sb->last_mount = (u32)ktime_get_real_seconds();
 	err = NULL;
 
 	get_page(bh->b_page);
@@ -701,7 +701,7 @@ static void bcache_device_detach(struct bcache_device *d)
 
 		SET_UUID_FLASH_ONLY(u, 0);
 		memcpy(u->uuid, invalid_uuid, 16);
-		u->invalidated = cpu_to_le32(get_seconds());
+		u->invalidated = cpu_to_le32((u32)ktime_get_real_seconds());
 		bch_uuid_write(d->c);
 	}
 
@@ -1027,7 +1027,7 @@ void bch_cached_dev_detach(struct cached_dev *dc)
 int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 			  uint8_t *set_uuid)
 {
-	uint32_t rtime = cpu_to_le32(get_seconds());
+	uint32_t rtime = cpu_to_le32((u32)ktime_get_real_seconds());
 	struct uuid_entry *u;
 	struct cached_dev *exist_dc, *t;
 
@@ -1070,7 +1070,7 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
 	    (BDEV_STATE(&dc->sb) == BDEV_STATE_STALE ||
 	     BDEV_STATE(&dc->sb) == BDEV_STATE_NONE)) {
 		memcpy(u->uuid, invalid_uuid, 16);
-		u->invalidated = cpu_to_le32(get_seconds());
+		u->invalidated = cpu_to_le32((u32)ktime_get_real_seconds());
 		u = NULL;
 	}
 
@@ -1390,7 +1390,7 @@ int bch_flash_dev_create(struct cache_set *c, uint64_t size)
 
 	get_random_bytes(u->uuid, 16);
 	memset(u->label, 0, 32);
-	u->first_reg = u->last_reg = cpu_to_le32(get_seconds());
+	u->first_reg = u->last_reg = cpu_to_le32((u32)ktime_get_real_seconds());
 
 	SET_UUID_FLASH_ONLY(u, 1);
 	u->sectors = size >> 9;
@@ -1894,7 +1894,7 @@ static void run_cache_set(struct cache_set *c)
 		goto err;
 
 	closure_sync(&cl);
-	c->sb.last_mount = get_seconds();
+	c->sb.last_mount = (u32)ktime_get_real_seconds();
 	bcache_write_super(c);
 
 	list_for_each_entry_safe(dc, t, &uncached_devices, list)
diff --git a/include/uapi/linux/bcache.h b/include/uapi/linux/bcache.h
index 821f71a2e48f..8d19e02d752a 100644
--- a/include/uapi/linux/bcache.h
+++ b/include/uapi/linux/bcache.h
@@ -195,7 +195,7 @@ struct cache_sb {
 	};
 	};
 
-	__u32			last_mount;	/* time_t */
+	__u32			last_mount;	/* time overflow in y2106 */
 
 	__u16			first_bucket;
 	union {
@@ -318,7 +318,7 @@ struct uuid_entry {
 		struct {
 			__u8	uuid[16];
 			__u8	label[32];
-			__u32	first_reg;
+			__u32	first_reg; /* time overflow in y2106 */
 			__u32	last_reg;
 			__u32	invalidated;
 
-- 
2.9.0

Re: [PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Coly Li]

On 2018/6/22 9:48 PM, Arnd Bergmann wrote:
> The get_seconds function is deprecated now since it returns a 32-bit
> value that will eventually overflow, and we are replacing it throughout
> the kernel with ktime_get_seconds() or ktime_get_real_seconds() that
> return a time64_t.
> 
> bcache uses get_seconds() to read the current system time and store it in
> the superblock as well as in uuid_entry structures that are user visible.
> 
> Unfortunately, the two structures in are still limited to 32 bits, so this
> won't fix any real problems but will still overflow in year 2106. Let's
> at least document that properly, in case we get an updated format in the
> future it can be fixed. We still have a long time before the overflow
> and checking the tools at https://github.com/koverstreet/bcache-tools
> reveals no access to any of them.
> 
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Added to 4.19 for-next.

Thanks.

Coly Li

> ---
>  drivers/md/bcache/super.c   | 12 ++++++------
>  include/uapi/linux/bcache.h |  4 ++--
>  2 files changed, 8 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/md/bcache/super.c b/drivers/md/bcache/super.c
> index fa4058e43202..74746d8ee05e 100644
> --- a/drivers/md/bcache/super.c
> +++ b/drivers/md/bcache/super.c
> @@ -181,7 +181,7 @@ static const char *read_super(struct cache_sb *sb, struct block_device *bdev,
>  		goto err;
>  	}
>  
> -	sb->last_mount = get_seconds();
> +	sb->last_mount = (u32)ktime_get_real_seconds();
>  	err = NULL;
>  
>  	get_page(bh->b_page);
> @@ -701,7 +701,7 @@ static void bcache_device_detach(struct bcache_device *d)
>  
>  		SET_UUID_FLASH_ONLY(u, 0);
>  		memcpy(u->uuid, invalid_uuid, 16);
> -		u->invalidated = cpu_to_le32(get_seconds());
> +		u->invalidated = cpu_to_le32((u32)ktime_get_real_seconds());
>  		bch_uuid_write(d->c);
>  	}
>  
> @@ -1027,7 +1027,7 @@ void bch_cached_dev_detach(struct cached_dev *dc)
>  int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
>  			  uint8_t *set_uuid)
>  {
> -	uint32_t rtime = cpu_to_le32(get_seconds());
> +	uint32_t rtime = cpu_to_le32((u32)ktime_get_real_seconds());
>  	struct uuid_entry *u;
>  	struct cached_dev *exist_dc, *t;
>  
> @@ -1070,7 +1070,7 @@ int bch_cached_dev_attach(struct cached_dev *dc, struct cache_set *c,
>  	    (BDEV_STATE(&dc->sb) == BDEV_STATE_STALE ||
>  	     BDEV_STATE(&dc->sb) == BDEV_STATE_NONE)) {
>  		memcpy(u->uuid, invalid_uuid, 16);
> -		u->invalidated = cpu_to_le32(get_seconds());
> +		u->invalidated = cpu_to_le32((u32)ktime_get_real_seconds());
>  		u = NULL;
>  	}
>  
> @@ -1390,7 +1390,7 @@ int bch_flash_dev_create(struct cache_set *c, uint64_t size)
>  
>  	get_random_bytes(u->uuid, 16);
>  	memset(u->label, 0, 32);
> -	u->first_reg = u->last_reg = cpu_to_le32(get_seconds());
> +	u->first_reg = u->last_reg = cpu_to_le32((u32)ktime_get_real_seconds());
>  
>  	SET_UUID_FLASH_ONLY(u, 1);
>  	u->sectors = size >> 9;
> @@ -1894,7 +1894,7 @@ static void run_cache_set(struct cache_set *c)
>  		goto err;
>  
>  	closure_sync(&cl);
> -	c->sb.last_mount = get_seconds();
> +	c->sb.last_mount = (u32)ktime_get_real_seconds();
>  	bcache_write_super(c);
>  
>  	list_for_each_entry_safe(dc, t, &uncached_devices, list)
> diff --git a/include/uapi/linux/bcache.h b/include/uapi/linux/bcache.h
> index 821f71a2e48f..8d19e02d752a 100644
> --- a/include/uapi/linux/bcache.h
> +++ b/include/uapi/linux/bcache.h
> @@ -195,7 +195,7 @@ struct cache_sb {
>  	};
>  	};
>  
> -	__u32			last_mount;	/* time_t */
> +	__u32			last_mount;	/* time overflow in y2106 */
>  
>  	__u16			first_bucket;
>  	union {
> @@ -318,7 +318,7 @@ struct uuid_entry {
>  		struct {
>  			__u8	uuid[16];
>  			__u8	label[32];
> -			__u32	first_reg;
> +			__u32	first_reg; /* time overflow in y2106 */
>  			__u32	last_reg;
>  			__u32	invalidated;
>  
> 

Re: [PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Arnd Bergmann]

On Sat, Jun 23, 2018 at 10:12 AM, Coly Li <colyli@suse.de> wrote:
> On 2018/6/22 9:48 PM, Arnd Bergmann wrote:
>> The get_seconds function is deprecated now since it returns a 32-bit
>> value that will eventually overflow, and we are replacing it throughout
>> the kernel with ktime_get_seconds() or ktime_get_real_seconds() that
>> return a time64_t.
>>
>> bcache uses get_seconds() to read the current system time and store it in
>> the superblock as well as in uuid_entry structures that are user visible.
>>
>> Unfortunately, the two structures in are still limited to 32 bits, so this
>> won't fix any real problems but will still overflow in year 2106. Let's
>> at least document that properly, in case we get an updated format in the
>> future it can be fixed. We still have a long time before the overflow
>> and checking the tools at https://github.com/koverstreet/bcache-tools
>> reveals no access to any of them.
>>
>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>
> Added to 4.19 for-next.

I noticed this isn't part of linux-next (yet). Did it get dropped?

      Arnd

Re: [PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Coly Li]

On 2018/7/13 10:20 PM, Arnd Bergmann wrote:
> On Sat, Jun 23, 2018 at 10:12 AM, Coly Li <colyli@suse.de> wrote:
>> On 2018/6/22 9:48 PM, Arnd Bergmann wrote:
>>> The get_seconds function is deprecated now since it returns a 32-bit
>>> value that will eventually overflow, and we are replacing it throughout
>>> the kernel with ktime_get_seconds() or ktime_get_real_seconds() that
>>> return a time64_t.
>>>
>>> bcache uses get_seconds() to read the current system time and store it in
>>> the superblock as well as in uuid_entry structures that are user visible.
>>>
>>> Unfortunately, the two structures in are still limited to 32 bits, so this
>>> won't fix any real problems but will still overflow in year 2106. Let's
>>> at least document that properly, in case we get an updated format in the
>>> future it can be fixed. We still have a long time before the overflow
>>> and checking the tools at https://github.com/koverstreet/bcache-tools
>>> reveals no access to any of them.
>>>
>>> Signed-off-by: Arnd Bergmann <arnd@arndb.de>
>>
>> Added to 4.19 for-next.
> 
> I noticed this isn't part of linux-next (yet). Did it get dropped?
> 
>       Arnd
> 
Hi Arnd,

It is here,
https://git.kernel.org/pub/scm/linux/kernel/git/colyli/bcache-patches.git/tree/for-next/0004-bcache-stop-using-the-deprecated-get_seconds.patch

It is for Linux v4.19, the merge window is not available yet, and I
don't post them to Jens yet.

Thanks.

Coly Li

Re: [PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Arnd Bergmann]

On Fri, Jul 13, 2018 at 5:34 PM, Coly Li <colyli@suse.de> wrote:
> On 2018/7/13 10:20 PM, Arnd Bergmann wrote:
>> On Sat, Jun 23, 2018 at 10:12 AM, Coly Li <colyli@suse.de> wrote:

>>> Added to 4.19 for-next.
>>
>> I noticed this isn't part of linux-next (yet). Did it get dropped?
>>
>>       Arnd
>>
> Hi Arnd,
>
> It is here,
> https://git.kernel.org/pub/scm/linux/kernel/git/colyli/bcache-patches.git/tree/for-next/0004-bcache-stop-using-the-deprecated-get_seconds.patch

Ok, thanks for the confirmation!

> It is for Linux v4.19, the merge window is not available yet, and I
> don't post them to Jens yet.

That sounds like an odd way to handle your patches. Normally all patches
should be part of linux-next *before* the merge window so they get tested
as part of the linux-next kernel before the pull request to Linus is sent.

It's not important for me for this particular patch, but generally speaking,
the Documentation/process/2.Process.rst states:

"all patches merged during a given merge window should really have found
their way into linux-next some time before the merge window opens."

      Arnd

Re: [PATCH] [v2] bcache: stop using the deprecated get_seconds()

[Jens Axboe]

On 7/13/18 12:57 PM, Arnd Bergmann wrote:
> On Fri, Jul 13, 2018 at 5:34 PM, Coly Li <colyli@suse.de> wrote:
>> On 2018/7/13 10:20 PM, Arnd Bergmann wrote:
>>> On Sat, Jun 23, 2018 at 10:12 AM, Coly Li <colyli@suse.de> wrote:
> 
>>>> Added to 4.19 for-next.
>>>
>>> I noticed this isn't part of linux-next (yet). Did it get dropped?
>>>
>>>       Arnd
>>>
>> Hi Arnd,
>>
>> It is here,
>> https://git.kernel.org/pub/scm/linux/kernel/git/colyli/bcache-patches.git/tree/for-next/0004-bcache-stop-using-the-deprecated-get_seconds.patch
> 
> Ok, thanks for the confirmation!
> 
>> It is for Linux v4.19, the merge window is not available yet, and I
>> don't post them to Jens yet.
> 
> That sounds like an odd way to handle your patches. Normally all patches
> should be part of linux-next *before* the merge window so they get tested
> as part of the linux-next kernel before the pull request to Linus is sent.
> 
> It's not important for me for this particular patch, but generally speaking,
> the Documentation/process/2.Process.rst states:
> 
> "all patches merged during a given merge window should really have found
> their way into linux-next some time before the merge window opens."

Coly does generally send before the merge window, I think the phrasing
is just a little bit off. Depending on risk, I usually tell folks to
submit their stuff for the next window at no later than -rc5 time,
so it'll be in for-next for a least two weeks before the merge window
opens.

-- 
Jens Axboe