plugins: Fix physical address calculation for IO regions

[PATCH] plugins: Fix physical address calculation for IO regions

Posted by Aaron Lindsay 2 years, 9 months ago

The address calculation for IO regions introduced by

commit 787148bf928a54b5cc86f5b434f9399e9737679c
Author: Aaron Lindsay <aaron@os.amperecomputing.com>
    plugins: Expose physical addresses instead of device offsets

is not always accurate. Use the more correct
MemoryRegionSection.offset_within_address_space.
---
 plugins/api.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/plugins/api.c b/plugins/api.c
index 5c1a413928..ba14e6f2b2 100644
--- a/plugins/api.c
+++ b/plugins/api.c
@@ -319,7 +319,7 @@ uint64_t qemu_plugin_hwaddr_phys_addr(const struct qemu_plugin_hwaddr *haddr)
             return block->offset + offset + block->mr->addr;
         } else {
             MemoryRegionSection *mrs = haddr->v.io.section;
-            return haddr->v.io.offset + mrs->mr->addr;
+            return mrs->offset_within_address_space + haddr->v.io.offset;
         }
     }
 #endif
-- 
2.17.1

Re: [PATCH] plugins: Fix physical address calculation for IO regions

Posted by Alex Bennée 2 years, 9 months ago

Aaron Lindsay <aaron@os.amperecomputing.com> writes:

> The address calculation for IO regions introduced by
>
> commit 787148bf928a54b5cc86f5b434f9399e9737679c
> Author: Aaron Lindsay <aaron@os.amperecomputing.com>
>     plugins: Expose physical addresses instead of device offsets

Queued to for-6.1/fixes-for-rc1, thanks.

>
> is not always accurate. Use the more correct
> MemoryRegionSection.offset_within_address_space.
> ---
>  plugins/api.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/plugins/api.c b/plugins/api.c
> index 5c1a413928..ba14e6f2b2 100644
> --- a/plugins/api.c
> +++ b/plugins/api.c
> @@ -319,7 +319,7 @@ uint64_t qemu_plugin_hwaddr_phys_addr(const struct qemu_plugin_hwaddr *haddr)
>              return block->offset + offset + block->mr->addr;
>          } else {
>              MemoryRegionSection *mrs = haddr->v.io.section;
> -            return haddr->v.io.offset + mrs->mr->addr;
> +            return mrs->offset_within_address_space + haddr->v.io.offset;
>          }
>      }
>  #endif


-- 
Alex Bennée

Re: [PATCH-for-6.1?] plugins: Fix physical address calculation for IO regions

Posted by Philippe Mathieu-Daudé 2 years, 9 months ago

On 7/20/21 9:57 PM, Aaron Lindsay wrote:
> The address calculation for IO regions introduced by
> 
> commit 787148bf928a54b5cc86f5b434f9399e9737679c
> Author: Aaron Lindsay <aaron@os.amperecomputing.com>
>     plugins: Expose physical addresses instead of device offsets
> 
> is not always accurate.

Indeed unlikely accurate in case of aliases (at least).

Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>

> Use the more correct
> MemoryRegionSection.offset_within_address_space.
> ---
>  plugins/api.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/plugins/api.c b/plugins/api.c
> index 5c1a413928..ba14e6f2b2 100644
> --- a/plugins/api.c
> +++ b/plugins/api.c
> @@ -319,7 +319,7 @@ uint64_t qemu_plugin_hwaddr_phys_addr(const struct qemu_plugin_hwaddr *haddr)
>              return block->offset + offset + block->mr->addr;
>          } else {
>              MemoryRegionSection *mrs = haddr->v.io.section;
> -            return haddr->v.io.offset + mrs->mr->addr;
> +            return mrs->offset_within_address_space + haddr->v.io.offset;
>          }
>      }
>  #endif
>