doc: DMA-mapping.txt has undeclared variables [Bug 10397]
PATCH v2 1/1 doc: DMA-mapping.txt has undeclared variables [Bug 10397]
From: Matti Linnanvuori <[email protected]>
Fix undeclared variables in code examples. Consistently use pdev as the variable of type struct pci_dev *. Bug 10397.
---
Documentation/DMA-mapping.txt +16 -16
@@ -315,9 +315,9 @@ dma_addr_t dma_handle; - cpu_addr = pci_alloc_consistent(dev, size, &dma_handle); + cpu_addr = pci_alloc_consistent(pdev, size, &dma_handle); -where dev is a struct pci_dev *. You should pass NULL for PCI like buses +where pdev is a struct pci_dev *. You should pass NULL for PCI like buses where devices don't have struct pci_dev (like ISA, EISA). This may be called in interrupt context.
@@ -354,9 +354,9 @@ To unmap and free such a DMA region, you call: - pci_free_consistent(dev, size, cpu_addr, dma_handle); + pci_free_consistent(pdev, size, cpu_addr, dma_handle); -where dev, size are the same as in the above call and cpu_addr and +where pdev, size are the same as in the above call and cpu_addr and dma_handle are the values pci_alloc_consistent returned to you. This function may not be called in interrupt context.
@@ -371,9 +371,9 @@ struct pci_pool *pool; - pool = pci_pool_create(name, dev, size, align, alloc); + pool = pci_pool_create(name, pdev, size, align, alloc); -The "name" is for diagnostics (like a kmem_cache name); dev and size +The "name" is for diagnostics (like a kmem_cache name); pdev and size are as above. The device's hardware alignment requirement for this type of data is "align" (which is expressed in bytes, and must be a power of two). If your device has no boundary crossing restrictions,
@@ -472,11 +472,11 @@ void *addr = buffer->ptr; size_t size = buffer->len; - dma_handle = pci_map_single(dev, addr, size, direction); + dma_handle = pci_map_single(pdev, addr, size, direction); and to unmap it: - pci_unmap_single(dev, dma_handle, size, direction); + pci_unmap_single(pdev, dma_handle, size, direction); You should call pci_unmap_single when the DMA activity is finished, e.g. from the interrupt which told you that the DMA transfer is done.
@@ -493,17 +493,17 @@ unsigned long offset = buffer->offset; size_t size = buffer->len; - dma_handle = pci_map_page(dev, page, offset, size, direction); + dma_handle = pci_map_page(pdev, page, offset, size, direction); ... - pci_unmap_page(dev, dma_handle, size, direction); + pci_unmap_page(pdev, dma_handle, size, direction); Here, "offset" means byte offset within the given page. With scatterlists, you map a region gathered from several regions by: - int i, count = pci_map_sg(dev, sglist, nents, direction); + int i, count = pci_map_sg(pdev, sglist, nents, direction); struct scatterlist *sg; for_each_sg(sglist, sg, count, i) {
@@ -527,7 +527,7 @@ To unmap a scatterlist, just call: - pci_unmap_sg(dev, sglist, nents, direction); + pci_unmap_sg(pdev, sglist, nents, direction); Again, make sure DMA activity has already finished.
@@ -550,11 +550,11 @@ So, firstly, just map it with pci_map_{single,sg}, and after each DMA transfer call either: - pci_dma_sync_single_for_cpu(dev, dma_handle, size, direction); + pci_dma_sync_single_for_cpu(pdev, dma_handle, size, direction); or: - pci_dma_sync_sg_for_cpu(dev, sglist, nents, direction); + pci_dma_sync_sg_for_cpu(pdev, sglist, nents, direction); as appropriate.
@@ -562,7 +562,7 @@ finish accessing the data with the cpu, and then before actually giving the buffer to the hardware call either: - pci_dma_sync_single_for_device(dev, dma_handle, size, direction); + pci_dma_sync_single_for_device(pdev, dma_handle, size, direction); or:
@@ -739,7 +739,7 @@ dma_addr_t dma_handle; - dma_handle = pci_map_single(dev, addr, size, direction); + dma_handle = pci_map_single(pdev, addr, size, direction); if (pci_dma_mapping_error(dma_handle)) { /* * reduce current DMA mapping usage,
____________________________________________________________________________________
You rock. That's why Blockbuster's offering you one month of Blockbuster Total Access, No Cost.
http://tc.deals.yahoo.com/tc/blockbuster/text5.com
Quoted 5 lines · Matti Linnanvuori
From: Matti Linnanvuori <[email protected]>
Fix undeclared variables in code examples. Consistently use pdev as the variable of type struct
pci_dev*. Bug 10397.
I think Grant's comment still stands, "fix undeclared variables" isn't an accurate description of what you're doing here...
That said, the patch itself looks fine to me, if a bit pedantic. The compiler will catch any mismatched arguments in this case, and the rest of the documentation is pretty clear that these routines take struct pci_dev not struct device, so I'd be inclined to just leave it the way it is to avoid churn.
Jesse
Quoted 9 lines · Jesse Barnes
Quoted 5 lines · Matti Linnanvuori
From: Matti Linnanvuori <[email protected]>
Fix undeclared variables in code examples. Consistently use pdev as the variable of type struct
pci_dev*. Bug 10397.I think Grant's comment still stands, "fix undeclared variables" isn't an accurate description of what you're doing here...
Yes, please just drop the first sentence. The second sentence is sufficient.
Quoted 5 lines
That said, the patch itself looks fine to me, if a bit pedantic. The compiler will catch any mismatched arguments in this case, and the rest of the documentation is pretty clear that these routines take struct
pci_devnot struct device, so I'd be inclined to just leave it the way it is to avoid churn.
I think it's a good habit to use "pdev" in device drivers and would like the documentation to incorporate that.
I'm ok with the churn since it's not conflicting with anything else AFAIK.
thanks,
grant
Quoted 20 lines · Grant Grundler
Quoted 9 lines · Jesse Barnes
Quoted 5 lines · Matti Linnanvuori
From: Matti Linnanvuori <[email protected]>
Fix undeclared variables in code examples. Consistently use pdev as the variable of type struct
pci_dev*. Bug 10397.I think Grant's comment still stands, "fix undeclared variables" isn't an accurate description of what you're doing here...
Yes, please just drop the first sentence. The second sentence is sufficient.
Quoted 5 lines
That said, the patch itself looks fine to me, if a bit pedantic. The compiler will catch any mismatched arguments in this case, and the rest of the documentation is pretty clear that these routines take struct
pci_devnot struct device, so I'd be inclined to just leave it the way it is to avoid churn.
Quoted 4 lines
I think it's a good habit to use "pdev" in device drivers and would like the documentation to incorporate that.
I'm ok with the churn since it's not conflicting with anything else AFAIK.
Matti, can you post an update with Grant's comments addressed? It's mainly about the changelog comment at this point, which should read something like "Update DMA mapping documentation to use 'pdev' rather than 'dev' in example code that calls routines expecting 'struct pci_device *', since 'dev' might make readers think they're passing 'struct device *' parameters."
Assuming it looks ok I'll go ahead and put it into the PCI tree.
Thanks,
Jesse