[Cbe-oss-dev] Context switching while page fault handling and warning

[Arnd Bergmann]

On Wednesday 14 June 2006 02:30, HyeonSeung Jang wrote:

>  
> static inline void restore_mfc_cntl(struct spu_state *csa, struct spu
> *spu)
> {
>              struct spu_priv2 __iomem *priv2 = spu->priv2;
>  
>              /* Restore, Step 72:
>               *    Restore the MFC_CNTL register for the CSA.
>               */
>              out_be64(&priv2->mfc_control_RW, csa-
> >priv2.mfc_control_RW);
>              eieio();
> }
>  
> After:
>  
> static inline void restore_mfc_cntl(struct spu_state *csa, struct spu
> *spu)
> {
>              struct spu_priv2 __iomem *priv2 = spu->priv2;
>  
>              /* Restore, Step 72:
>               *    Restore the MFC_CNTL register for the CSA.
>               */
>              out_be64(&priv2->mfc_control_RW, csa-
> >priv2.mfc_control_RW);
>              eieio();
>              if( (csa->priv2.mfc_control_RW &
> MFC_CNTL_SUSPEND_DMA_QUEUE_MASK)) {
>                            out_be64(&priv2->mfc_control_RW,
> MFC_CNTL_RESTART_DMA_COMMAND);
>                            eieio();
>              }
> }
>  
> It seems to work.
>  
>  
>  
> Now I have two questions :
>  
> 1) My fix is correct ? or are there any side effects of it ?

The fix looks technically correct, but I don't think its optimal.
Instead we should better have the context save/restore code
remember whether the context was about to handle a fault
and then process that later.

I'm adding your fix to the next release for now, but mark it
as something we need to work on again later.

> 2) Why there is following warning on mambo simulator? :
>    'WARNING: 61392752237: SPE2: MFC_CMD_QUEUE channel count of 15 is
> inconsistent with number of available DMA queue entries of 16'
>  
That's an unrelated but known problem we still need to debug.

	Arnd <><