[U-Boot] [PATCH v3] linux/compat.h: port lower_32_bits and upper_32_bits from Linux

[backport from linux commit 204b885e and 218e180e7] 64 bit processors are becomming more and more popular. lower_32_bits and upper_32_bits save our labor doing shifts/manipulations like (u32)(n) and (u32)((n) >> 32). They are good helpers in both little and big endian cases. Port these two functions here from Linux:include/linux/kernel.h, cater the comment message to little/big endian cases, also remove the definition in drivers/usb/host/xhci.h. Later on, developers could include linux/compat.h if they want to use these two functions.
Signed-off-by: Lijun Pan Lijun.Pan@freescale.com --- v3: change the comment message for lower_32_bits() and upper_32_bits() change the commit message to argue the necessity of this patch. v2: add git SHA of linux kernel tree.
drivers/usb/host/xhci.h | 4 +--- include/linux/compat.h | 19 +++++++++++++++++++ 2 files changed, 20 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/host/xhci.h b/drivers/usb/host/xhci.h index ceb1573..6381596 100644 --- a/drivers/usb/host/xhci.h +++ b/drivers/usb/host/xhci.h @@ -20,9 +20,7 @@ #include <asm/cache.h> #include <asm/io.h> #include <linux/list.h> - -#define upper_32_bits(n) (u32)((n) >> 32) -#define lower_32_bits(n) (u32)(n) +#include <linux/compat.h>
#define MAX_EP_CTX_NUM 31 #define XHCI_ALIGNMENT 64 diff --git a/include/linux/compat.h b/include/linux/compat.h index 3fdfb39..35e216e 100644 --- a/include/linux/compat.h +++ b/include/linux/compat.h @@ -57,4 +57,23 @@ , __FILE__, __LINE__); }
#define PAGE_SIZE 4096 + +/** + * upper_32_bits - return MSB bits 32-63 of a number if little endian, or + * return MSB bits 0-31 of a number if big endian. + * @n: the number we're accessing + * + * A basic shift-right of a 64- or 32-bit quantity. Use this to suppress + * the "right shift count >= width of type" warning when that quantity is + * 32-bits. + */ +#define upper_32_bits(n) ((u32)(((n) >> 16) >> 16)) + +/** + * lower_32_bits - return LSB bits 0-31 of a number if little endian, or + * return LSB bits 32-63 of a number if big endian. + * @n: the number we're accessing + */ +#define lower_32_bits(n) ((u32)(n)) + #endif

Dear Lijun Pan,
In message 1402690280-8848-1-git-send-email-Lijun.Pan@freescale.com you wrote:
[backport from linux commit 204b885e and 218e180e7] 64 bit processors are becomming more and more popular. lower_32_bits and upper_32_bits save our labor doing shifts/manipulations like (u32)(n) and (u32)((n) >> 32). They are good helpers in both little and big endian cases. Port these two functions here from Linux:include/linux/kernel.h, cater the comment message to little/big endian cases, also remove the definition in drivers/usb/host/xhci.h. Later on, developers could include linux/compat.h if they want to use these two functions.
Signed-off-by: Lijun Pan Lijun.Pan@freescale.com
v3: change the comment message for lower_32_bits() and upper_32_bits() change the commit message to argue the necessity of this patch. v2: add git SHA of linux kernel tree.
I think it would be better to split this commit into two separate patches.
One, which in my opinion should be part of a patch series that adds more users of these functions, would add the new functions. The other would then convert the only existing user (drivers/usb/host/xhci.h) to use these.
But before we convert drivers/usb/host/xhci.h, we should fix the FIXME there - this looks very much like a bug to me:
1119 static inline void xhci_writeq(__le64 volatile *regs, const u64 val) 1120 { 1121 __u32 *ptr = (__u32 *)regs; 1122 u32 val_lo = lower_32_bits(val); 1123 /* FIXME */ 1124 u32 val_hi = 0;
I think this should rather be
u32 val_hi = upper_32_bits(val);
and I cannot think of any good reason why it is not...
Best regards,
Wolfgang Denk
participants (2)
-
Lijun Pan
-
Wolfgang Denk