[PATCH] drm/xe: avoid plain 64-bit division

[PATCH] drm/xe: avoid plain 64-bit division

[Arnd Bergmann]

From: Arnd Bergmann <arnd@arndb.de>

Building the xe driver for i386 results in a link time warning:

x86_64-linux-ld: drivers/gpu/drm/xe/xe_migrate.o: in function `xe_migrate_vram':
xe_migrate.c:(.text+0x1e15): undefined reference to `__udivdi3'

Avoid this by using DIV_U64_ROUND_UP() instead of DIV_ROUND_UP().  The driver
is unlikely to be used on 32=bit hardware, so the extra cost here is not
too important.

Fixes: 9c44fd5f6e8a ("drm/xe: Add migrate layer functions for SVM support")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 drivers/gpu/drm/xe/xe_migrate.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/xe/xe_migrate.c b/drivers/gpu/drm/xe/xe_migrate.c
index df4282c71bf0..aafb6209f2d0 100644
--- a/drivers/gpu/drm/xe/xe_migrate.c
+++ b/drivers/gpu/drm/xe/xe_migrate.c
@@ -1547,7 +1547,7 @@ void xe_migrate_wait(struct xe_migrate *m)
 static u32 pte_update_cmd_size(u64 size)
 {
 	u32 num_dword;
-	u64 entries = DIV_ROUND_UP(size, XE_PAGE_SIZE);
+	u64 entries = DIV_U64_ROUND_UP(size, XE_PAGE_SIZE);
 
 	XE_WARN_ON(size > MAX_PREEMPTDISABLE_TRANSFER);
 	/*
@@ -1558,7 +1558,7 @@ static u32 pte_update_cmd_size(u64 size)
 	 * 2 dword for the page table's physical location
 	 * 2*n dword for value of pte to fill (each pte entry is 2 dwords)
 	 */
-	num_dword = (1 + 2) * DIV_ROUND_UP(entries, 0x1ff);
+	num_dword = (1 + 2) * DIV_U64_ROUND_UP(entries, 0x1ff);
 	num_dword += entries * 2;
 
 	return num_dword;
-- 
2.39.5

Re: [PATCH] drm/xe: avoid plain 64-bit division

[Rodrigo Vivi]

On Mon, Mar 24, 2025 at 10:06:02PM +0100, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> Building the xe driver for i386 results in a link time warning:
> 
> x86_64-linux-ld: drivers/gpu/drm/xe/xe_migrate.o: in function `xe_migrate_vram':
> xe_migrate.c:(.text+0x1e15): undefined reference to `__udivdi3'
> 
> Avoid this by using DIV_U64_ROUND_UP() instead of DIV_ROUND_UP().  The driver
> is unlikely to be used on 32=bit hardware, so the extra cost here is not
> too important.
> 
> Fixes: 9c44fd5f6e8a ("drm/xe: Add migrate layer functions for SVM support")
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Reviewed-by: Rodrigo Vivi <rodrigo.vivi@intel.com>

and pushing it right now, thanks for the patch

> ---
>  drivers/gpu/drm/xe/xe_migrate.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/gpu/drm/xe/xe_migrate.c b/drivers/gpu/drm/xe/xe_migrate.c
> index df4282c71bf0..aafb6209f2d0 100644
> --- a/drivers/gpu/drm/xe/xe_migrate.c
> +++ b/drivers/gpu/drm/xe/xe_migrate.c
> @@ -1547,7 +1547,7 @@ void xe_migrate_wait(struct xe_migrate *m)
>  static u32 pte_update_cmd_size(u64 size)
>  {
>  	u32 num_dword;
> -	u64 entries = DIV_ROUND_UP(size, XE_PAGE_SIZE);
> +	u64 entries = DIV_U64_ROUND_UP(size, XE_PAGE_SIZE);
>  
>  	XE_WARN_ON(size > MAX_PREEMPTDISABLE_TRANSFER);
>  	/*
> @@ -1558,7 +1558,7 @@ static u32 pte_update_cmd_size(u64 size)
>  	 * 2 dword for the page table's physical location
>  	 * 2*n dword for value of pte to fill (each pte entry is 2 dwords)
>  	 */
> -	num_dword = (1 + 2) * DIV_ROUND_UP(entries, 0x1ff);
> +	num_dword = (1 + 2) * DIV_U64_ROUND_UP(entries, 0x1ff);
>  	num_dword += entries * 2;
>  
>  	return num_dword;
> -- 
> 2.39.5
>