[Qemu-devel] [PATCH] hw/cadence_gem: Make rx_desc_addr and tx_desc_addr uint32_t

[Qemu-devel] [PATCH] hw/cadence_gem: Make rx_desc_addr and tx_desc_addr uint32_t

[Peter Maydell]

Make the state fields rx_desc_addr and tx_desc_addr uint32_t;
this matches the VMStateDescription, and also conforms to how
hardware works: the registers don't magically become larger
if the device is attached to a CPU with a larger physical
address size. It also fixes a compile failure if the
target_phys_addr_t type is changed to 64 bits.

Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
I'm going through fixing problems with moving target-arm to
a larger physical address width so we can support the A15
Large Physical Address Extensions...

 hw/cadence_gem.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/hw/cadence_gem.c b/hw/cadence_gem.c
index e2140ae..e563409 100644
--- a/hw/cadence_gem.c
+++ b/hw/cadence_gem.c
@@ -339,8 +339,8 @@ typedef struct {
     uint8_t phy_loop; /* Are we in phy loopback? */
 
     /* The current DMA descriptor pointers */
-    target_phys_addr_t rx_desc_addr;
-    target_phys_addr_t tx_desc_addr;
+    uint32_t rx_desc_addr;
+    uint32_t tx_desc_addr;
 
 } GemState;
 
-- 
1.7.1

Re: [Qemu-devel] [PATCH] hw/cadence_gem: Make rx_desc_addr and tx_desc_addr uint32_t

[Peter Maydell]

Peter C: ping? I was hoping for a review or ack for this
one before I stick it into an arm-devs pullreq.

thanks
-- PMM

On 22 May 2012 18:02, Peter Maydell <peter.maydell@linaro.org> wrote:
> Make the state fields rx_desc_addr and tx_desc_addr uint32_t;
> this matches the VMStateDescription, and also conforms to how
> hardware works: the registers don't magically become larger
> if the device is attached to a CPU with a larger physical
> address size. It also fixes a compile failure if the
> target_phys_addr_t type is changed to 64 bits.
>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
> ---
> I'm going through fixing problems with moving target-arm to
> a larger physical address width so we can support the A15
> Large Physical Address Extensions...
>
>  hw/cadence_gem.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/hw/cadence_gem.c b/hw/cadence_gem.c
> index e2140ae..e563409 100644
> --- a/hw/cadence_gem.c
> +++ b/hw/cadence_gem.c
> @@ -339,8 +339,8 @@ typedef struct {
>     uint8_t phy_loop; /* Are we in phy loopback? */
>
>     /* The current DMA descriptor pointers */
> -    target_phys_addr_t rx_desc_addr;
> -    target_phys_addr_t tx_desc_addr;
> +    uint32_t rx_desc_addr;
> +    uint32_t tx_desc_addr;
>
>  } GemState;
>
> --
> 1.7.1

Re: [Qemu-devel] [PATCH] hw/cadence_gem: Make rx_desc_addr and tx_desc_addr uint32_t

[Peter Crosthwaite]

On Wed, May 23, 2012 at 3:02 AM, Peter Maydell <peter.maydell@linaro.org> wrote:
> Make the state fields rx_desc_addr and tx_desc_addr uint32_t;
> this matches the VMStateDescription, and also conforms to how
> hardware works: the registers don't magically become larger
> if the device is attached to a CPU with a larger physical
> address size. It also fixes a compile failure if the
> target_phys_addr_t type is changed to 64 bits.
>
> Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Peter A. G. Crosthwaite <peter.crosthwaite@petalogix.com>

> ---
> I'm going through fixing problems with moving target-arm to
> a larger physical address width so we can support the A15
> Large Physical Address Extensions...
>
>  hw/cadence_gem.c |    4 ++--
>  1 files changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/hw/cadence_gem.c b/hw/cadence_gem.c
> index e2140ae..e563409 100644
> --- a/hw/cadence_gem.c
> +++ b/hw/cadence_gem.c
> @@ -339,8 +339,8 @@ typedef struct {
>     uint8_t phy_loop; /* Are we in phy loopback? */
>
>     /* The current DMA descriptor pointers */
> -    target_phys_addr_t rx_desc_addr;
> -    target_phys_addr_t tx_desc_addr;
> +    uint32_t rx_desc_addr;
> +    uint32_t tx_desc_addr;
>
>  } GemState;
>
> --
> 1.7.1
>