]> git.baikalelectronics.ru Git - kernel.git/commit
defxx: Correct the receive DMA map size
authorMaciej W. Rozycki <macro@linux-mips.org>
Sat, 5 Jul 2014 14:14:22 +0000 (15:14 +0100)
committerDavid S. Miller <davem@davemloft.net>
Tue, 8 Jul 2014 22:30:10 +0000 (15:30 -0700)
commit151e9f7bde8123ca667c03f2c0fe37730df60829
tree65d6ad13d4d7a4b2de150eb84280a2e6ba674563
parent848aa49431c4995b0ae258053373d4076aa12fb6
defxx: Correct the receive DMA map size

Receive DMA maps are oversized, they include EISA legacy 128-byte
alignment padding in size calculation whereas this padding is never used
for data.  Worse yet, if the skb's data area has been realigned indeed,
then data beyond the end of the buffer will be synchronised from the
receive DMA bounce buffer, possibly corrupting data structures residing
in memory beyond the actual end of this data buffer.

Therefore switch to using PI_RCV_DATA_K_SIZE_MAX rather than NEW_SKB_SIZE
in DMA mapping, the value the former macro expands to is written to the
receive ring DMA descriptor of the PDQ DMA chip and determines the
maximum amount of data PDQ will ever transfer to the corresponding data
buffer, including all headers and padding.

Reported-by: Robert Coerver <Robert.Coerver@ll.mit.edu>
Tested-by: Robert Coerver <Robert.Coerver@ll.mit.edu>
Signed-off-by: Maciej W. Rozycki <macro@linux-mips.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
drivers/net/fddi/defxx.c