[PATCH v2] Btrfs: use do_div to avoid compile errors on 32bit box

[PATCH v2] Btrfs: use do_div to avoid compile errors on 32bit box

[Liu Bo]

When doing div operation of u64 type, we need to be careful and use do_div
to avoid compile ERROR on 32bit box:

"ERROR: "__udivdi3" [fs/btrfs/btrfs.ko] undefined!
make[1]: *** [__modpost] Error 1

v1->v2:
- fix stupid do_div() with type "signed integer".

Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
---
 fs/btrfs/extent-tree.c |   11 ++++++-----
 1 files changed, 6 insertions(+), 5 deletions(-)

diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
index 80d6148..e43e4f1 100644
--- a/fs/btrfs/extent-tree.c
+++ b/fs/btrfs/extent-tree.c
@@ -6735,9 +6735,10 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)
 	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
 	struct btrfs_device *device;
 	u64 min_free;
+	u32 dev_min = 1;
+	u32 dev_nr = 0;
+	u32 dup = 2;
 	int index;
-	int dev_nr = 0;
-	int dev_min = 1;
 	int full = 0;
 	int ret = 0;
 
@@ -6796,14 +6797,14 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)
 	index = get_block_group_index(block_group);
 	if (index == 0) {
 		dev_min = 4;
-		min_free /= 2;
+		do_div(min_free, dup);
 	} else if (index == 1) {
 		dev_min = 2;
 	} else if (index == 2) {
-		min_free *= 2;
+		min_free *= dup;
 	} else if (index == 3) {
 		dev_min = fs_devices->rw_devices;
-		min_free /= dev_min;
+		do_div(min_free, dev_min);
 	}
 
 	mutex_lock(&root->fs_info->chunk_mutex);
-- 
1.6.5.2

Re: [PATCH v2] Btrfs: use do_div to avoid compile errors on 32bit box

[Josef Bacik]

On 08/19/2011 10:17 PM, Liu Bo wrote:
> When doing div operation of u64 type, we need to be careful and use do_div
> to avoid compile ERROR on 32bit box:
> 
> "ERROR: "__udivdi3" [fs/btrfs/btrfs.ko] undefined!
> make[1]: *** [__modpost] Error 1
> 
> v1->v2:
> - fix stupid do_div() with type "signed integer".
> 
> Signed-off-by: Liu Bo <liubo2009@cn.fujitsu.com>
> ---
>  fs/btrfs/extent-tree.c |   11 ++++++-----
>  1 files changed, 6 insertions(+), 5 deletions(-)
> 
> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c
> index 80d6148..e43e4f1 100644
> --- a/fs/btrfs/extent-tree.c
> +++ b/fs/btrfs/extent-tree.c
> @@ -6735,9 +6735,10 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)
>  	struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices;
>  	struct btrfs_device *device;
>  	u64 min_free;
> +	u32 dev_min = 1;
> +	u32 dev_nr = 0;
> +	u32 dup = 2;
>  	int index;
> -	int dev_nr = 0;
> -	int dev_min = 1;
>  	int full = 0;
>  	int ret = 0;
>  
> @@ -6796,14 +6797,14 @@ int btrfs_can_relocate(struct btrfs_root *root, u64 bytenr)
>  	index = get_block_group_index(block_group);
>  	if (index == 0) {
>  		dev_min = 4;
> -		min_free /= 2;
> +		do_div(min_free, dup);

I think Linus was less complaining about how you're dividing here and
more about the fact that you are.  A divide by 2 is the same as a >> 1.
 I'll send a patch to fix this.  Thanks,

Josef

Re: [PATCH v2] Btrfs: use do_div to avoid compile errors on 32bit box

[Linus Torvalds]

On Sat, Aug 20, 2011 at 5:21 AM, Josef Bacik <josef@redhat.com> wrote:
>
> I think Linus was less complaining about how you're dividing here and
> more about the fact that you are. =A0A divide by 2 is the same as a >=
> 1.
> =A0I'll send a patch to fix this. =A0Thanks,

Indeed. A single-bit 64-bit "double shift" may be a few cycles, but
it's still pretty damn cheap. Especially when compared to a 64x32 bit
divide. We absolutely don't want to do "do_div()" in order to divide
by the constant 2.

                     Linus