The legacy API wrappers in include/linux/pci-dma-compat.h
should go away as it creates unnecessary midlayering
for include/linux/dma-mapping.h APIs, instead use dma-mapping.h
APIs directly.
The patch has been generated with the coccinelle script below
and compile-tested.
@@@@
- PCI_DMA_BIDIRECTIONAL
+ DMA_BIDIRECTIONAL
@@@@
- PCI_DMA_TODEVICE
+ DMA_TO_DEVICE
@@@@
- PCI_DMA_FROMDEVICE
+ DMA_FROM_DEVICE
@@@@
- PCI_DMA_NONE
+ DMA_NONE
@@ expression E1, E2, E3; @@
- pci_alloc_consistent(E1, E2, E3)
+ dma_alloc_coherent(&E1->dev, E2, E3, GFP_ATOMIC)
@@ expression E1, E2, E3; @@
- pci_zalloc_consistent(E1, E2, E3)
+ dma_alloc_coherent(&E1->dev, E2, E3, GFP_ATOMIC)
@@ expression E1, E2, E3, E4; @@
- pci_free_consistent(E1, E2, E3, E4)
+ dma_free_coherent(&E1->dev, E2, E3, E4)
@@ expression E1, E2, E3, E4; @@
- pci_map_single(E1, E2, E3, E4)
+ dma_map_single(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_unmap_single(E1, E2, E3, E4)
+ dma_unmap_single(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4, E5; @@
- pci_map_page(E1, E2, E3, E4, E5)
+ dma_map_page(&E1->dev, E2, E3, E4, (enum dma_data_direction)E5)
@@ expression E1, E2, E3, E4; @@
- pci_unmap_page(E1, E2, E3, E4)
+ dma_unmap_page(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_map_sg(E1, E2, E3, E4)
+ dma_map_sg(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_unmap_sg(E1, E2, E3, E4)
+ dma_unmap_sg(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_dma_sync_single_for_cpu(E1, E2, E3, E4)
+ dma_sync_single_for_cpu(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_dma_sync_single_for_device(E1, E2, E3, E4)
+ dma_sync_single_for_device(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_dma_sync_sg_for_cpu(E1, E2, E3, E4)
+ dma_sync_sg_for_cpu(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2, E3, E4; @@
- pci_dma_sync_sg_for_device(E1, E2, E3, E4)
+ dma_sync_sg_for_device(&E1->dev, E2, E3, (enum dma_data_direction)E4)
@@ expression E1, E2; @@
- pci_dma_mapping_error(E1, E2)
+ dma_mapping_error(&E1->dev, E2)
@@ expression E1, E2; @@
- pci_set_consistent_dma_mask(E1, E2)
+ dma_set_coherent_mask(&E1->dev, E2)
@@ expression E1, E2; @@
- pci_set_dma_mask(E1, E2)
+ dma_set_mask(&E1->dev, E2)
Signed-off-by: Suraj Upadhyay <[email protected]>
---
This change is proposed by Christoph Hellwig <[email protected]>
in the post https://marc.info/?l=kernel-janitors&m=158745678307186&w=4
on kernel-janitors Mailing List.
drivers/staging/qlge/qlge_mpi.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/drivers/staging/qlge/qlge_mpi.c b/drivers/staging/qlge/qlge_mpi.c
index fa178fc642a6..16a9bf818346 100644
--- a/drivers/staging/qlge/qlge_mpi.c
+++ b/drivers/staging/qlge/qlge_mpi.c
@@ -788,8 +788,9 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
char *my_buf;
dma_addr_t buf_dma;
- my_buf = pci_alloc_consistent(qdev->pdev, word_count * sizeof(u32),
- &buf_dma);
+ my_buf = dma_alloc_coherent(&qdev->pdev->dev,
+ word_count * sizeof(u32), &buf_dma,
+ GFP_ATOMIC);
if (!my_buf)
return -EIO;
@@ -797,8 +798,8 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
if (!status)
memcpy(buf, my_buf, word_count * sizeof(u32));
- pci_free_consistent(qdev->pdev, word_count * sizeof(u32), my_buf,
- buf_dma);
+ dma_free_coherent(&qdev->pdev->dev, word_count * sizeof(u32), my_buf,
+ buf_dma);
return status;
}
--
2.17.1
On 2020-07-11 18:16 +0530, Suraj Upadhyay wrote:
> The legacy API wrappers in include/linux/pci-dma-compat.h
> should go away as it creates unnecessary midlayering
> for include/linux/dma-mapping.h APIs, instead use dma-mapping.h
> APIs directly.
>
> The patch has been generated with the coccinelle script below
> and compile-tested.
>
[...]
>
> @@ expression E1, E2, E3, E4; @@
> - pci_dma_sync_single_for_device(E1, E2, E3, E4)
> + dma_sync_single_for_device(&E1->dev, E2, E3, (enum dma_data_direction)E4)
The qlge driver contains more usages of the deprecated pci_dma_* api
than what this diff addresses. In particular, there are some calls to
pci_dma_sync_single_for_cpu() which were not changed despite this
expression being in the semantic patch.
Dunno what happened but it should be reviewed. After converting away
from all of the old api, the TODO file should also be updated.
[...]
>
> diff --git a/drivers/staging/qlge/qlge_mpi.c b/drivers/staging/qlge/qlge_mpi.c
> index fa178fc642a6..16a9bf818346 100644
> --- a/drivers/staging/qlge/qlge_mpi.c
> +++ b/drivers/staging/qlge/qlge_mpi.c
> @@ -788,8 +788,9 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
> char *my_buf;
> dma_addr_t buf_dma;
>
> - my_buf = pci_alloc_consistent(qdev->pdev, word_count * sizeof(u32),
> - &buf_dma);
> + my_buf = dma_alloc_coherent(&qdev->pdev->dev,
> + word_count * sizeof(u32), &buf_dma,
> + GFP_ATOMIC);
> if (!my_buf)
> return -EIO;
>
> @@ -797,8 +798,8 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
> if (!status)
> memcpy(buf, my_buf, word_count * sizeof(u32));
>
> - pci_free_consistent(qdev->pdev, word_count * sizeof(u32), my_buf,
> - buf_dma);
> + dma_free_coherent(&qdev->pdev->dev, word_count * sizeof(u32), my_buf,
> + buf_dma);
> return status;
> }
>
> --
> 2.17.1
>
On Mon, Jul 13, 2020 at 01:59:59PM +0900, Benjamin Poirier wrote:
> On 2020-07-11 18:16 +0530, Suraj Upadhyay wrote:
> > The legacy API wrappers in include/linux/pci-dma-compat.h
> > should go away as it creates unnecessary midlayering
> > for include/linux/dma-mapping.h APIs, instead use dma-mapping.h
> > APIs directly.
> >
> > The patch has been generated with the coccinelle script below
> > and compile-tested.
> >
> [...]
> >
> > @@ expression E1, E2, E3, E4; @@
> > - pci_dma_sync_single_for_device(E1, E2, E3, E4)
> > + dma_sync_single_for_device(&E1->dev, E2, E3, (enum dma_data_direction)E4)
>
> The qlge driver contains more usages of the deprecated pci_dma_* api
> than what this diff addresses. In particular, there are some calls to
> pci_dma_sync_single_for_cpu() which were not changed despite this
> expression being in the semantic patch.
Hii Ben,
I couldn't find any instances of pci_dma_sync_single_for_cpu in
the drivers/staging/qlge/ driver, I ran a simple `git grep pci_dma_sync_single_for_cpu/device`
and got nothing.
If I am wrong, please send the line number of the usages.
> Dunno what happened but it should be reviewed. After converting away
> from all of the old api, the TODO file should also be updated.
Thanks for reminding me this, I would send a follow up patch to remove
"pci_dma_*" from "avoid legacy/deprecated apis (ex. replace pci_dma_*, replace pci_enable_msi,
use pci_iomap)".
Thanks and Cheers,
Suraj Upadhyay.
> [...]
>
> >
> > diff --git a/drivers/staging/qlge/qlge_mpi.c b/drivers/staging/qlge/qlge_mpi.c
> > index fa178fc642a6..16a9bf818346 100644
> > --- a/drivers/staging/qlge/qlge_mpi.c
> > +++ b/drivers/staging/qlge/qlge_mpi.c
> > @@ -788,8 +788,9 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
> > char *my_buf;
> > dma_addr_t buf_dma;
> >
> > - my_buf = pci_alloc_consistent(qdev->pdev, word_count * sizeof(u32),
> > - &buf_dma);
> > + my_buf = dma_alloc_coherent(&qdev->pdev->dev,
> > + word_count * sizeof(u32), &buf_dma,
> > + GFP_ATOMIC);
> > if (!my_buf)
> > return -EIO;
> >
> > @@ -797,8 +798,8 @@ int ql_dump_risc_ram_area(struct ql_adapter *qdev, void *buf,
> > if (!status)
> > memcpy(buf, my_buf, word_count * sizeof(u32));
> >
> > - pci_free_consistent(qdev->pdev, word_count * sizeof(u32), my_buf,
> > - buf_dma);
> > + dma_free_coherent(&qdev->pdev->dev, word_count * sizeof(u32), my_buf,
> > + buf_dma);
> > return status;
> > }
> >
> > --
> > 2.17.1
> >
>
>
On 2020-07-13 11:14 +0530, Suraj Upadhyay wrote:
> On Mon, Jul 13, 2020 at 01:59:59PM +0900, Benjamin Poirier wrote:
> > On 2020-07-11 18:16 +0530, Suraj Upadhyay wrote:
> > > The legacy API wrappers in include/linux/pci-dma-compat.h
> > > should go away as it creates unnecessary midlayering
> > > for include/linux/dma-mapping.h APIs, instead use dma-mapping.h
> > > APIs directly.
> > >
> > > The patch has been generated with the coccinelle script below
> > > and compile-tested.
> > >
> > [...]
> > >
> > > @@ expression E1, E2, E3, E4; @@
> > > - pci_dma_sync_single_for_device(E1, E2, E3, E4)
> > > + dma_sync_single_for_device(&E1->dev, E2, E3, (enum dma_data_direction)E4)
> >
> > The qlge driver contains more usages of the deprecated pci_dma_* api
> > than what this diff addresses. In particular, there are some calls to
> > pci_dma_sync_single_for_cpu() which were not changed despite this
> > expression being in the semantic patch.
>
> Hii Ben,
> I couldn't find any instances of pci_dma_sync_single_for_cpu in
> the drivers/staging/qlge/ driver, I ran a simple `git grep pci_dma_sync_single_for_cpu/device`
> and got nothing.
> If I am wrong, please send the line number of the usages.
You're right, sorry. I was missing commit e955a071b9b3 ("staging: qlge:
replace deprecated apis pci_dma_*") in my tree.