Re: Strange patch to the Z85230 driver.

Re: Strange patch to the Z85230 driver.

[William Lee Irwin III]

On Sat, Oct 12, 2002 at 07:02:29PM +0100, Alan Cox wrote:
> These are DMA ring buffers for ISA DMA, they do not need to be zeroed. 

get_free_page() was #defined to get_zeroed_page() to begin with.

/mnt/b/2.5.38/linux-2.5/include/linux/gfp.h:83:#define get_free_page get_zeroed_page


Bill

Strange patch to the Z85230 driver.

[Alan Cox]

These are DMA ring buffers for ISA DMA, they do not need to be zeroed. 

diff -Nru a/drivers/net/wan/z85230.c b/drivers/net/wan/z85230.c
--- a/drivers/net/wan/z85230.c	Fri Oct 11 21:22:51 2002
+++ b/drivers/net/wan/z85230.c	Fri Oct 11 21:22:51 2002
@@ -889,12 +889,12 @@
 	if(c->mtu  > PAGE_SIZE/2)
 		return -EMSGSIZE;
 	 
-	c->rx_buf[0]=(void *)get_free_page(GFP_KERNEL|GFP_DMA);
+	c->rx_buf[0]=(void *)get_zeroed_page(GFP_KERNEL|GFP_DMA);
 	if(c->rx_buf[0]==NULL)
 		return -ENOBUFS;
 	c->rx_buf[1]=c->rx_buf[0]+PAGE_SIZE/2;
 	
-	c->tx_dma_buf[0]=(void *)get_free_page(GFP_KERNEL|GFP_DMA);
+	c->tx_dma_buf[0]=(void *)get_zeroed_page(GFP_KERNEL|GFP_DMA);
 	if(c->tx_dma_buf[0]==NULL)
 	{
 		free_page((unsigned long)c->rx_buf[0]);
@@ -1079,7 +1079,7 @@
 	if(c->mtu  > PAGE_SIZE/2)
 		return -EMSGSIZE;
 	 
-	c->tx_dma_buf[0]=(void *)get_free_page(GFP_KERNEL|GFP_DMA);
+	c->tx_dma_buf[0]=(void *)get_zeroed_page(GFP_KERNEL|GFP_DMA);
 	if(c->tx_dma_buf[0]==NULL)
 		return -ENOBUFS;
 

Re: Strange patch to the Z85230 driver.

[Linus Torvalds]

On 12 Oct 2002, Alan Cox wrote:
>
> These are DMA ring buffers for ISA DMA, they do not need to be zeroed. 

Note that the code is the same as it alwasy was - the patch is part of a 
rename: "get_free_page()" is gone and is now called "get_zeroed_page()" 
so that people see what it does.

"get_free_page()" has always zeroed the contents, with "__get_free_page()" 
being the old traditional (and now long non-zeroing variety.

So the patch is purely cosmetic, and if the code was strange before, it's 
strange now.

And it's perhaps more _clearly_ strange, which was really the whole point
of the patch.

See the part of the changeset that touches <linux/gfp.h>:

	--- a/include/linux/gfp.h       Sat Oct 12 11:16:59 2002
	+++ b/include/linux/gfp.h       Sat Oct 12 11:16:59 2002
	@@ -71,11 +71,6 @@
	                __get_free_pages((gfp_mask) | GFP_DMA,(order))
	 
	 /*
	- * The old interface name will be removed in 2.5:
	- */
	-#define get_free_page get_zeroed_page
	-
	-/*
	  * There is only one 'core' page-freeing function.
	  */
	 extern void FASTCALL(__free_pages(struct page *page, unsigned int order));

to see that even the comment said this should be done.

		Linus

Re: Strange patch to the Z85230 driver.

[Alan Cox]

On Sat, 2002-10-12 at 19:19, Linus Torvalds wrote:
> So the patch is purely cosmetic, and if the code was strange before, it's 
> strange now.
> 
> And it's perhaps more _clearly_ strange, which was really the whole point
> of the patch.

Well it certainly worked 8)