Re: AW: PowerPC PCI DMA issues (prefetch/coherency?)

Hi,

With the default config for the Sequoia board on 2.6.24, calling 
pci_dma_sync_sg_for_cpu() results in executing
invalidate_dcache_range() in arch/ppc/kernel/misc.S from __dma_sync().  
This OOPses on PPC440 since it tries to call directly the assembly 
instruction dcbi, which can only be executed in supervisor mode.  We 
tried that before resorting to manual cache line management with 
usermode-safe assembly calls.

Regards,
Tom Burns
International Datacasting Corporation

Mikhail Zolotaryov wrote:

Hi,

Why manage cache lines  manually, if appropriate code is a part of 
__dma_sync / dma_sync_single_for_device of DMA API ? (implies 
CONFIG_NOT_COHERENT_CACHE enabled, as default for Sequoia Board)

Prodyut Hazarika wrote:

quoted

Hi Adam,

 

quoted

Yes, I am using the 440EPx (same as the sequoia board). Our 
ideDriver is DMA'ing blocks of 192-byte data over the PCI bus
    

(using
 

quoted

the Sil0680A PCI-IDE bridge). Most of the DMA's (depending on timing)
end up being partially corrupted when we try to parse the data in the
virtual page. We have confirmed the data is good before the PCI-IDE
bridge. We are creating two 8K pages and map them to physical DMA
    

memory
 

quoted

using single-entry scatter/gather structs. When a DMA block is
corrupted, we see a random portion of it (always a multiple of 16byte
cache lines) is overwritten with old data from the last time the
    

buffer
 

quoted

was used.     

This looks like a cache coherency problem.
Can you ensure that the TLB entries corresponding to the DMA region has
the CacheInhibit bit set.
You will need a BDI connected to your system.

Also, you will need to invalidate and flush the lines appropriately,
since in 440 cores,
L1Cache coherency is managed entirely by software.
Please look at drivers/net/ibm_newemac/mal.c and core.c for example on
how to do it.

Thanks
Prodyut

On Thu, 2009-09-03 at 13:27 -0700, Prodyut Hazarika wrote:
 

quoted

Hi Adam,

   

quoted

Are you sure there is L2 cache on the 440?
      

It depends on the SoC you are using. SoC like 460EX (Canyonlands
    

board)
 

quoted

have L2Cache.
It seems you are using a Sequoia board, which has a 440EPx SoC. 440EPx
has a 440 cpu core, but no L2Cache.
Could you please tell me which SoC you are using?
You can also refer to the appropriate dts file to see if there is L2C.
For example, in canyonlands.dts (460EX based board), we have the L2C
entry.
        L2C0: l2c {
              ...
        }

   

quoted

I am seeing this problem with our custom IDE driver which is based on
      

quoted

quoted

pretty old code. Our driver uses pci_alloc_consistent() to allocate
      

the
 

quoted

quoted

physical DMA memory and alloc_pages() to allocate a virtual page. 
It then uses pci_map_sg() to map to a scatter/gather buffer. 
Perhaps I should convert these to the DMA API calls as you suggest.
      

Could you give more details on the consistency problem? It is a good
idea to change to the new DMA APIs, but pci_alloc_consistent() should
work too

Thanks
Prodyut   

On Thu, 2009-09-03 at 19:57 +1000, Benjamin Herrenschmidt wrote:
   

quoted

On Thu, 2009-09-03 at 09:05 +0100, Chris Pringle wrote:
     

quoted

Hi Adam,

If you have a look in include/asm-ppc/pgtable.h for the following
        

section:
   

quoted

quoted

#ifdef CONFIG_44x
#define _PAGE_BASE    (_PAGE_PRESENT | _PAGE_ACCESSED |
        

_PAGE_GUARDED)
   

quoted

quoted

#else
#define _PAGE_BASE    (_PAGE_PRESENT | _PAGE_ACCESSED)
#endif

Try adding _PAGE_COHERENT to the appropriate line above and see if
        

that    

quoted

quoted

fixes your issue - this causes the 'M' bit to be set on the page
        

which    

quoted

quoted

sure enforce cache coherency. If it doesn't, you'll need to check
        

the    

quoted

quoted

'M' bit isn't being masked out in head_44x.S (it was originally
        

masked    

quoted

quoted

out on arch/powerpc, but was fixed in later kernels when the cache
        

quoted

quoted

quoted

coherency issues with non-SMP systems were resolved).
        

I have some doubts about the usefulness of doing that for 4xx.
      

AFAIK,
 

quoted

quoted

the 440 core just ignores M.

The problem lies probably elsewhere. Maybe the L2 cache coherency
      

isn't
   

quoted

enabled or not working ?

The L1 cache on 440 is simply not coherent, so drivers have to make
      

sure
   

quoted

they use the appropriate DMA APIs which will do cache flushing when
needed.

Adam, what driver is causing you that sort of problems ?

Cheers,
Ben.