Re: [PATCH 20/27] Initconst section fixes for SCSI

[James Bottomley]

On Thu, 2011-09-15 at 13:09 -0700, Andi Kleen wrote:
> From: Andi Kleen <ak@xxxxxxxxxxxxxxx>

I don't know why you're doing this without a description (and please cc
at least linux-scsi on SCSI patches).

> Cc: JBottomley@xxxxxxxxxxxxx
> Signed-off-by: Andi Kleen <ak@xxxxxxxxxxxxxxx>
> ---
>  drivers/scsi/aacraid/linit.c        |    2 +-
>  drivers/scsi/aic94xx/aic94xx_init.c |    2 +-
>  2 files changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/scsi/aacraid/linit.c b/drivers/scsi/aacraid/linit.c
> index 3382475..9c8d000 100644
> --- a/drivers/scsi/aacraid/linit.c
> +++ b/drivers/scsi/aacraid/linit.c
> @@ -92,7 +92,7 @@ static DECLARE_PCI_DEVICE_TABLE(aac_pci_tbl) = {
>  #elif defined(__devinitconst)
>  static const struct pci_device_id aac_pci_tbl[] __devinitconst = {
>  #else
> -static const struct pci_device_id aac_pci_tbl[] __devinitdata = {
> +static const struct pci_device_id aac_pci_tbl[] __devinitconst = {

How can this possibly be correct?  You just checked in the preceding
#elif defined() that __devinitconst wasn't defined for this leg, so if
it's taken there'll be an instant compile failure.

If you want to argue the whole of the #ifdef construct should be removed
then I'd be interested.

James

èº{.nÇ+‰·Ÿ®‰­†+%ŠËlzwm…ébëæìr¸›zX§»®w¥Š{ayºÊÚë,j­¢f£¢·hš‹àz¹®w¥¢¸¢·¦j:+v‰¨ŠwèjØm¶Ÿÿ¾«‘êçzZ+ƒùšŽŠÝj"ú!¶iO•æ¬z·švØ^¶m§ÿðÃnÆàþY&—