Skip to contentSkip to conversation

pci_dma_set_mask()

RFC Zach Brown 3 participants 5 messages Feb 28, 2001

lore
Share & tools
lore

Zach Brown

author
#0 lore raw

This patch really has two parts. Most of it adds a helper function that does the

	if(pci_dma_supported()) { ->dma_mask = mask }

code path. I was using the api today and didn't realize that I had to
set the mask myself, I assumed the _supported call would do it. If
people prefer the struct setting api, thats fine with me :)

It also seems goofy that pci_dma_supported() unconditionally return true on x86. If a device needs a mask smaller than GFP_DMA (admitedly probably won't exist in the real world, one hopes), they're going to have troubles because x86 just falls back to GFP_DMA when the mask isn't 0xffffffff..

all totally untested, of course..

Signature · 58 lines
 zach

--- linux-2.4.2/include/linux/pci.h.dmasup	Wed Feb 28 10:26:14 2001
+++ linux-2.4.2/include/linux/pci.h	Wed Feb 28 10:30:12 2001
@@ -527,6 +527,7 @@
 
 int pci_enable_device(struct pci_dev *dev);
 void pci_set_master(struct pci_dev *dev);
+int pci_set_dma_mask(struct pci_dev *dev, dma_addr_t mask);
 int pci_set_power_state(struct pci_dev *dev, int state);
 int pci_assign_resource(struct pci_dev *dev, int i);
 
--- linux-2.4.2/include/asm-i386/pci.h.dmasup	Wed Feb 28 10:19:01 2001
+++ linux-2.4.2/include/asm-i386/pci.h	Wed Feb 28 10:25:40 2001
@@ -152,6 +152,14 @@
  */
 extern inline int pci_dma_supported(struct pci_dev *hwdev, dma_addr_t mask)
 {
+        /*
+         * we fall back to GFP_DMA when the mask isn't all 1s,
+         * so we can't guarantee allocations that must be
+         * within a tighter range than GFP_DMA..
+         */
+        if(mask < 0x00ffffff)
+                return 0;
+
 	return 1;
 }
 
--- linux-2.4.2/drivers/pci/pci.c.dmasup	Wed Feb 28 10:26:34 2001
+++ linux-2.4.2/drivers/pci/pci.c	Wed Feb 28 10:27:34 2001
@@ -518,6 +518,18 @@
 	pcibios_set_master(dev);
 }
 
+int
+pci_set_dma_mask(struct pci_dev *dev, dma_addr_t mask)
+{
+    if(! pci_dma_supported(dev, mask))
+        return 0;
+
+    dev->dma_mask = mask;
+
+    return 1;
+}
+    
+
 /*
  * Translate the low bits of the PCI base
  * to the resource type
@@ -1206,6 +1218,7 @@
 EXPORT_SYMBOL(pci_find_slot);
 EXPORT_SYMBOL(pci_find_subsys);
 EXPORT_SYMBOL(pci_set_master);
+EXPORT_SYMBOL(pci_set_dma_mask);
 EXPORT_SYMBOL(pci_set_power_state);
 EXPORT_SYMBOL(pci_assign_resource);
 EXPORT_SYMBOL(pci_register_driver);
#1 lore raw

Review

Quoted 10 lines · Zach Brown
(quoted diff) +10 -0
+int
+pci_set_dma_mask(struct pci_dev *dev, dma_addr_t mask)
+{
+    if(! pci_dma_supported(dev, mask))
+        return 0;
+
+    dev->dma_mask = mask;
+
+    return 1;
+}

pci_dma_supported has a boolean return, but the kernel norm is to return zero on success, and -EFOO on error. I like your proposal with the extremely minor nit that I think pci_set_dma_mask should return ENODEV or EIO or something on error, and zero on success.

Jeff

Signature · 3 lines
Jeff Garzik       | "You see, in this world there's two kinds of
Building 1024     |  people, my friend: Those with loaded guns
MandrakeSoft      |  and those who dig. You dig."  --Blondie
author
#2 lore raw
In reply to Zach Brown
Quoted 2 lines

pci_dma_supported has a boolean return, but the kernel norm is to return zero on success, and -EFOO on error. I like your proposal with the

*nod* I just followed pci_dma_supported().

Quoted 2 lines

extremely minor nit that I think pci_set_dma_mask should return ENODEV or EIO or something on error, and zero on success.

I agree, though I'd like to leave the decision up to people who live and breathe this stuff.

please feel free to make minor adjustments and submit :)

- z

#3 lore raw
Quoted 7 lines · Zach Brown
Quoted 2 lines

extremely minor nit that I think pci_set_dma_mask should return ENODEV or EIO or something on error, and zero on success.

I agree, though I'd like to leave the decision up to people who live and breathe this stuff.

please feel free to make minor adjustments and submit :)

Jeff/Zach, I agree, I'm fully for such a patch, but please update the documentation! It is the most important part of the patch.

Later,
David S. Miller
[email protected]

author
#4 lore raw
In reply to Zach Brown
Quoted 2 lines · David S. Miller

Jeff/Zach, I agree, I'm fully for such a patch, but please update the documentation! It is the most important part of the patch.

Very good point. I'll add Jeff's error returning and spin some minor docs and resend.

thanks.