mbox series

[0/3] pci: Let PCI DMA API functions propagate a MemTxResult

Message ID 20200904162620.657726-1-philmd@redhat.com
Headers show
Series pci: Let PCI DMA API functions propagate a MemTxResult | expand

Message

Philippe Mathieu-Daudé Sept. 4, 2020, 4:26 p.m. UTC
The DMA API propagates MemTxResult:
- MEMTX_OK,
- MEMTX_device_ERROR,
- MEMTX_DECODE_ERROR.

Let the PCI DMA API propagate them, as they are
clearer than an undocumented 'int'.

Based-on: <20200904154439.643272-1-philmd@redhat.com>
https://lists.gnu.org/archive/html/qemu-devel/2020-09/msg02048.html

Philippe Mathieu-Daudé (3):
  pci: Let pci_dma_rw() propagate MemTxResult
  pci: Let pci_dma_read() propagate MemTxResult
  pci: Let pci_dma_write() propagate MemTxResult

 include/hw/pci/pci.h | 50 ++++++++++++++++++++++++++++++++++++++------
 1 file changed, 44 insertions(+), 6 deletions(-)

Comments

Li Qiang Sept. 5, 2020, 2:41 a.m. UTC | #1
Philippe Mathieu-Daudé <philmd@redhat.com> 于2020年9月5日周六 上午12:26写道:
>
> dma_memory_rw() returns a MemTxResult type.
> Do not discard it, return it to the caller.
>
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
>  include/hw/pci/pci.h | 18 ++++++++++++++++--
>  1 file changed, 16 insertions(+), 2 deletions(-)
>
> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> index a221dfb3b08..a85b0bc3c44 100644
> --- a/include/hw/pci/pci.h
> +++ b/include/hw/pci/pci.h
> @@ -785,8 +785,22 @@ static inline AddressSpace *pci_get_address_space(PCIDevice *dev)
>      return &dev->bus_master_as;
>  }
>
> -static inline int pci_dma_rw(PCIDevice *dev, dma_addr_t addr,
> -                             void *buf, dma_addr_t len, DMADirection dir)
> +/**
> + * pci_dma_rw: Read from or write to an address space from PCI device.
> + *
> + * Return a MemTxResult indicating whether the operation succeeded
> + * or failed (eg unassigned memory, device rejected the transaction,
> + * IOMMU fault).
> + *
> + * @dev: #PCIDevice doing the memory access
> + * @addr: address within the #PCIDevice address space
> + * @buf: buffer with the data transferred
> + * @len: the number of bytes to read or write
> + * @dir: indicates the transfer direction
> + */
> +static inline MemTxResult pci_dma_rw(PCIDevice *dev, dma_addr_t addr,
> +                                     void *buf, dma_addr_t len,
> +                                     DMADirection dir)
>  {
>      return dma_memory_rw(pci_get_address_space(dev), addr, buf, len,
>                           dir, MEMTXATTRS_UNSPECIFIED);

Reviewed-by: Li Qiang <liq3ea@gmail.com>

> --
> 2.26.2
>
>
Li Qiang Sept. 5, 2020, 2:42 a.m. UTC | #2
Philippe Mathieu-Daudé <philmd@redhat.com> 于2020年9月5日周六 上午12:26写道:
>
> pci_dma_rw() returns a MemTxResult type.
> Do not discard it, return it to the caller.
>
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>

Reviewed-by: Li Qiang <liq3ea@gmail.com>

> ---
>  include/hw/pci/pci.h | 16 ++++++++++++++--
>  1 file changed, 14 insertions(+), 2 deletions(-)
>
> diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h
> index bb57525dffd..76826e240a9 100644
> --- a/include/hw/pci/pci.h
> +++ b/include/hw/pci/pci.h
> @@ -824,8 +824,20 @@ static inline MemTxResult pci_dma_read(PCIDevice *dev, dma_addr_t addr,
>      return pci_dma_rw(dev, addr, buf, len, DMA_DIRECTION_TO_DEVICE);
>  }
>
> -static inline int pci_dma_write(PCIDevice *dev, dma_addr_t addr,
> -                                const void *buf, dma_addr_t len)
> +/**
> + * pci_dma_write: Write to address space from PCI device.
> + *
> + * Return a MemTxResult indicating whether the operation succeeded
> + * or failed (eg unassigned memory, device rejected the transaction,
> + * IOMMU fault).
> + *
> + * @dev: #PCIDevice doing the memory access
> + * @addr: address within the #PCIDevice address space
> + * @buf: buffer with the data transferred
> + * @len: the number of bytes to write
> + */
> +static inline MemTxResult pci_dma_write(PCIDevice *dev, dma_addr_t addr,
> +                                        const void *buf, dma_addr_t len)
>  {
>      return pci_dma_rw(dev, addr, (void *) buf, len, DMA_DIRECTION_FROM_DEVICE);
>  }
> --
> 2.26.2
>
>
Richard Henderson Sept. 6, 2020, 4:08 a.m. UTC | #3
On 9/4/20 9:26 AM, Philippe Mathieu-Daudé wrote:
> pci_dma_rw() returns a MemTxResult type.
> Do not discard it, return it to the caller.
> 
> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
> ---
>  include/hw/pci/pci.h | 16 ++++++++++++++--
>  1 file changed, 14 insertions(+), 2 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~
Philippe Mathieu-Daudé Sept. 15, 2020, 5:30 p.m. UTC | #4
On 9/4/20 6:26 PM, Philippe Mathieu-Daudé wrote:
> The DMA API propagates MemTxResult:
> - MEMTX_OK,
> - MEMTX_device_ERROR,
> - MEMTX_DECODE_ERROR.
> 
> Let the PCI DMA API propagate them, as they are
> clearer than an undocumented 'int'.
> 
> Based-on: <20200904154439.643272-1-philmd@redhat.com>
> https://lists.gnu.org/archive/html/qemu-devel/2020-09/msg02048.html
> 
> Philippe Mathieu-Daudé (3):
>   pci: Let pci_dma_rw() propagate MemTxResult
>   pci: Let pci_dma_read() propagate MemTxResult
>   pci: Let pci_dma_write() propagate MemTxResult
> 
>  include/hw/pci/pci.h | 50 ++++++++++++++++++++++++++++++++++++++------
>  1 file changed, 44 insertions(+), 6 deletions(-)
> 

This series is fully review.

Paolo, if you take the DMA series on which this one
is based, could you take this too?

Thanks :)