From: Chris Leech <christopher.leech@intel.com>
Date: Thu, 8 Mar 2007 17:57:33 +0000 (-0800)
Subject: ioatdma: Push pending transactions to hardware more frequently
X-Git-Url: https://git.karo-electronics.de/?a=commitdiff_plain;h=000725d56a196e72dc22328324c5ec5506265736;p=linux-beck.git

ioatdma: Push pending transactions to hardware more frequently
Every 20 descriptors turns out to be to few append commands with
newer/faster CPUs.  Pushing every 4 still cuts down on MMIO writes to an
acceptable level without letting the DMA engine run out of work.

Signed-off-by: Chris Leech <christopher.leech@intel.com>
---

diff --git a/drivers/dma/ioatdma.c b/drivers/dma/ioatdma.c
index 850014139556..4e6afd752034 100644
--- a/drivers/dma/ioatdma.c
+++ b/drivers/dma/ioatdma.c
@@ -310,7 +310,7 @@ static dma_cookie_t do_ioat_dma_memcpy(struct ioat_dma_chan *ioat_chan,
 	list_splice_init(&new_chain, ioat_chan->used_desc.prev);
 
 	ioat_chan->pending += desc_count;
-	if (ioat_chan->pending >= 20) {
+	if (ioat_chan->pending >= 4) {
 		append = 1;
 		ioat_chan->pending = 0;
 	}
@@ -818,7 +818,7 @@ static void __devexit ioat_remove(struct pci_dev *pdev)
 }
 
 /* MODULE API */
-MODULE_VERSION("1.7");
+MODULE_VERSION("1.9");
 MODULE_LICENSE("GPL");
 MODULE_AUTHOR("Intel Corporation");