Re: [PATCH v4 1/2] memremap: rename MEMORY_DEVICE_DEVDAX to MEMORY_DEVICE_GENERIC
On Mon, Aug 31, 2020 at 12:19:07PM +0200, Roger Pau Monné wrote:
> On Thu, Aug 20, 2020 at 01:37:41PM +0200, Roger Pau Monné wrote:
> > On Tue, Aug 11, 2020 at 11:07:36PM +0200, David Hildenbrand wrote:
> > > On 11.08.20 11:44, Roger Pau Monne wrote:
> > > > This is in preparation for the logic behind MEMORY_DEVICE_DEVDAX also
> > > > being used by non DAX devices.
> > > >
> > > > No functional change intended.
> > > >
> > > > Signed-off-by: Roger Pau Monné
Dan is out on leave so I'll chime in.
I can't really justify keeping this as DEVDAX if there is another user who
needs the same type of mapping.
Sorry for the delay.
Reviewed-by: Ira Weiny
> > > > ---
> > > > Cc: Dan Williams
> > > > Cc: Vishal Verma
> > > > Cc: Dave Jiang
> > > > Cc: Andrew Morton
> > > > Cc: Jason Gunthorpe
> > > > Cc: Ira Weiny
> > > > Cc: "Aneesh Kumar K.V"
> > > > Cc: Johannes Thumshirn
> > > > Cc: Logan Gunthorpe
> > > > Cc: linux-nvd...@lists.01.org
> > > > Cc: xen-de...@lists.xenproject.org
> > > > Cc: linux...@kvack.org
> > > > ---
> > > > drivers/dax/device.c | 2 +-
> > > > include/linux/memremap.h | 9 -
> > > > mm/memremap.c| 2 +-
> > > > 3 files changed, 6 insertions(+), 7 deletions(-)
> > > >
> > > > diff --git a/drivers/dax/device.c b/drivers/dax/device.c
> > > > index 4c0af2eb7e19..1e89513f3c59 100644
> > > > --- a/drivers/dax/device.c
> > > > +++ b/drivers/dax/device.c
> > > > @@ -429,7 +429,7 @@ int dev_dax_probe(struct device *dev)
> > > > return -EBUSY;
> > > > }
> > > >
> > > > - dev_dax->pgmap.type = MEMORY_DEVICE_DEVDAX;
> > > > + dev_dax->pgmap.type = MEMORY_DEVICE_GENERIC;
> > > > addr = devm_memremap_pages(dev, _dax->pgmap);
> > > > if (IS_ERR(addr))
> > > > return PTR_ERR(addr);
> > > > diff --git a/include/linux/memremap.h b/include/linux/memremap.h
> > > > index 5f5b2df06e61..e5862746751b 100644
> > > > --- a/include/linux/memremap.h
> > > > +++ b/include/linux/memremap.h
> > > > @@ -46,11 +46,10 @@ struct vmem_altmap {
> > > > * wakeup is used to coordinate physical address space management (ex:
> > > > * fs truncate/hole punch) vs pinned pages (ex: device dma).
> > > > *
> > > > - * MEMORY_DEVICE_DEVDAX:
> > > > + * MEMORY_DEVICE_GENERIC:
> > > > * Host memory that has similar access semantics as System RAM i.e. DMA
> > > > - * coherent and supports page pinning. In contrast to
> > > > - * MEMORY_DEVICE_FS_DAX, this memory is access via a device-dax
> > > > - * character device.
> > > > + * coherent and supports page pinning. This is for example used by DAX
> > > > devices
> > > > + * that expose memory using a character device.
> > > > *
> > > > * MEMORY_DEVICE_PCI_P2PDMA:
> > > > * Device memory residing in a PCI BAR intended for use with
> > > > Peer-to-Peer
> > > > @@ -60,7 +59,7 @@ enum memory_type {
> > > > /* 0 is reserved to catch uninitialized type fields */
> > > > MEMORY_DEVICE_PRIVATE = 1,
> > > > MEMORY_DEVICE_FS_DAX,
> > > > - MEMORY_DEVICE_DEVDAX,
> > > > + MEMORY_DEVICE_GENERIC,
> > > > MEMORY_DEVICE_PCI_P2PDMA,
> > > > };
> > > >
> > > > diff --git a/mm/memremap.c b/mm/memremap.c
> > > > index 03e38b7a38f1..006dace60b1a 100644
> > > > --- a/mm/memremap.c
> > > > +++ b/mm/memremap.c
> > > > @@ -216,7 +216,7 @@ void *memremap_pages(struct dev_pagemap *pgmap, int
> > > > nid)
> > > > return ERR_PTR(-EINVAL);
> > > > }
> > > > break;
> > > > - case MEMORY_DEVICE_DEVDAX:
> > > > + case MEMORY_DEVICE_GENERIC:
> > > > need_devmap_managed = false;
> > > > break;
> > > > case MEMORY_DEVICE_PCI_P2PDMA:
> > > >
> > >
> > > No strong opinion (@Dan?), I do wonder if a separate type would make
> > > sense.
> >
> > Gentle ping.
>
> Sorry to ping again, but I would rather get this out of my queue if
> possible, seeing as the other patch is OK to go in but depends on this
> one going in first.
>
> Thanks, Roger.
Re: [PATCH v4 1/2] memremap: rename MEMORY_DEVICE_DEVDAX to MEMORY_DEVICE_GENERIC
On Tue, 11 Aug 2020 11:44:46 +0200 Roger Pau Monne wrote: > This is in preparation for the logic behind MEMORY_DEVICE_DEVDAX also > being used by non DAX devices. Acked-by: Andrew Morton . Please add it to the Xen tree when appropriate. (I'm not sure what David means by "separate type", but we can do that later if desired. Dan is taking a taking a bit of downtime).
Re: [PATCH v4 1/2] memremap: rename MEMORY_DEVICE_DEVDAX to MEMORY_DEVICE_GENERIC
On Thu, Aug 20, 2020 at 01:37:41PM +0200, Roger Pau Monné wrote:
> On Tue, Aug 11, 2020 at 11:07:36PM +0200, David Hildenbrand wrote:
> > On 11.08.20 11:44, Roger Pau Monne wrote:
> > > This is in preparation for the logic behind MEMORY_DEVICE_DEVDAX also
> > > being used by non DAX devices.
> > >
> > > No functional change intended.
> > >
> > > Signed-off-by: Roger Pau Monné
> > > ---
> > > Cc: Dan Williams
> > > Cc: Vishal Verma
> > > Cc: Dave Jiang
> > > Cc: Andrew Morton
> > > Cc: Jason Gunthorpe
> > > Cc: Ira Weiny
> > > Cc: "Aneesh Kumar K.V"
> > > Cc: Johannes Thumshirn
> > > Cc: Logan Gunthorpe
> > > Cc: linux-nvd...@lists.01.org
> > > Cc: xen-de...@lists.xenproject.org
> > > Cc: linux...@kvack.org
> > > ---
> > > drivers/dax/device.c | 2 +-
> > > include/linux/memremap.h | 9 -
> > > mm/memremap.c| 2 +-
> > > 3 files changed, 6 insertions(+), 7 deletions(-)
> > >
> > > diff --git a/drivers/dax/device.c b/drivers/dax/device.c
> > > index 4c0af2eb7e19..1e89513f3c59 100644
> > > --- a/drivers/dax/device.c
> > > +++ b/drivers/dax/device.c
> > > @@ -429,7 +429,7 @@ int dev_dax_probe(struct device *dev)
> > > return -EBUSY;
> > > }
> > >
> > > - dev_dax->pgmap.type = MEMORY_DEVICE_DEVDAX;
> > > + dev_dax->pgmap.type = MEMORY_DEVICE_GENERIC;
> > > addr = devm_memremap_pages(dev, _dax->pgmap);
> > > if (IS_ERR(addr))
> > > return PTR_ERR(addr);
> > > diff --git a/include/linux/memremap.h b/include/linux/memremap.h
> > > index 5f5b2df06e61..e5862746751b 100644
> > > --- a/include/linux/memremap.h
> > > +++ b/include/linux/memremap.h
> > > @@ -46,11 +46,10 @@ struct vmem_altmap {
> > > * wakeup is used to coordinate physical address space management (ex:
> > > * fs truncate/hole punch) vs pinned pages (ex: device dma).
> > > *
> > > - * MEMORY_DEVICE_DEVDAX:
> > > + * MEMORY_DEVICE_GENERIC:
> > > * Host memory that has similar access semantics as System RAM i.e. DMA
> > > - * coherent and supports page pinning. In contrast to
> > > - * MEMORY_DEVICE_FS_DAX, this memory is access via a device-dax
> > > - * character device.
> > > + * coherent and supports page pinning. This is for example used by DAX
> > > devices
> > > + * that expose memory using a character device.
> > > *
> > > * MEMORY_DEVICE_PCI_P2PDMA:
> > > * Device memory residing in a PCI BAR intended for use with Peer-to-Peer
> > > @@ -60,7 +59,7 @@ enum memory_type {
> > > /* 0 is reserved to catch uninitialized type fields */
> > > MEMORY_DEVICE_PRIVATE = 1,
> > > MEMORY_DEVICE_FS_DAX,
> > > - MEMORY_DEVICE_DEVDAX,
> > > + MEMORY_DEVICE_GENERIC,
> > > MEMORY_DEVICE_PCI_P2PDMA,
> > > };
> > >
> > > diff --git a/mm/memremap.c b/mm/memremap.c
> > > index 03e38b7a38f1..006dace60b1a 100644
> > > --- a/mm/memremap.c
> > > +++ b/mm/memremap.c
> > > @@ -216,7 +216,7 @@ void *memremap_pages(struct dev_pagemap *pgmap, int
> > > nid)
> > > return ERR_PTR(-EINVAL);
> > > }
> > > break;
> > > - case MEMORY_DEVICE_DEVDAX:
> > > + case MEMORY_DEVICE_GENERIC:
> > > need_devmap_managed = false;
> > > break;
> > > case MEMORY_DEVICE_PCI_P2PDMA:
> > >
> >
> > No strong opinion (@Dan?), I do wonder if a separate type would make sense.
>
> Gentle ping.
Sorry to ping again, but I would rather get this out of my queue if
possible, seeing as the other patch is OK to go in but depends on this
one going in first.
Thanks, Roger.
Re: [PATCH v4 1/2] memremap: rename MEMORY_DEVICE_DEVDAX to MEMORY_DEVICE_GENERIC
On 11.08.20 11:44, Roger Pau Monne wrote:
> This is in preparation for the logic behind MEMORY_DEVICE_DEVDAX also
> being used by non DAX devices.
>
> No functional change intended.
>
> Signed-off-by: Roger Pau Monné
> ---
> Cc: Dan Williams
> Cc: Vishal Verma
> Cc: Dave Jiang
> Cc: Andrew Morton
> Cc: Jason Gunthorpe
> Cc: Ira Weiny
> Cc: "Aneesh Kumar K.V"
> Cc: Johannes Thumshirn
> Cc: Logan Gunthorpe
> Cc: linux-nvd...@lists.01.org
> Cc: xen-de...@lists.xenproject.org
> Cc: linux...@kvack.org
> ---
> drivers/dax/device.c | 2 +-
> include/linux/memremap.h | 9 -
> mm/memremap.c| 2 +-
> 3 files changed, 6 insertions(+), 7 deletions(-)
>
> diff --git a/drivers/dax/device.c b/drivers/dax/device.c
> index 4c0af2eb7e19..1e89513f3c59 100644
> --- a/drivers/dax/device.c
> +++ b/drivers/dax/device.c
> @@ -429,7 +429,7 @@ int dev_dax_probe(struct device *dev)
> return -EBUSY;
> }
>
> - dev_dax->pgmap.type = MEMORY_DEVICE_DEVDAX;
> + dev_dax->pgmap.type = MEMORY_DEVICE_GENERIC;
> addr = devm_memremap_pages(dev, _dax->pgmap);
> if (IS_ERR(addr))
> return PTR_ERR(addr);
> diff --git a/include/linux/memremap.h b/include/linux/memremap.h
> index 5f5b2df06e61..e5862746751b 100644
> --- a/include/linux/memremap.h
> +++ b/include/linux/memremap.h
> @@ -46,11 +46,10 @@ struct vmem_altmap {
> * wakeup is used to coordinate physical address space management (ex:
> * fs truncate/hole punch) vs pinned pages (ex: device dma).
> *
> - * MEMORY_DEVICE_DEVDAX:
> + * MEMORY_DEVICE_GENERIC:
> * Host memory that has similar access semantics as System RAM i.e. DMA
> - * coherent and supports page pinning. In contrast to
> - * MEMORY_DEVICE_FS_DAX, this memory is access via a device-dax
> - * character device.
> + * coherent and supports page pinning. This is for example used by DAX
> devices
> + * that expose memory using a character device.
> *
> * MEMORY_DEVICE_PCI_P2PDMA:
> * Device memory residing in a PCI BAR intended for use with Peer-to-Peer
> @@ -60,7 +59,7 @@ enum memory_type {
> /* 0 is reserved to catch uninitialized type fields */
> MEMORY_DEVICE_PRIVATE = 1,
> MEMORY_DEVICE_FS_DAX,
> - MEMORY_DEVICE_DEVDAX,
> + MEMORY_DEVICE_GENERIC,
> MEMORY_DEVICE_PCI_P2PDMA,
> };
>
> diff --git a/mm/memremap.c b/mm/memremap.c
> index 03e38b7a38f1..006dace60b1a 100644
> --- a/mm/memremap.c
> +++ b/mm/memremap.c
> @@ -216,7 +216,7 @@ void *memremap_pages(struct dev_pagemap *pgmap, int nid)
> return ERR_PTR(-EINVAL);
> }
> break;
> - case MEMORY_DEVICE_DEVDAX:
> + case MEMORY_DEVICE_GENERIC:
> need_devmap_managed = false;
> break;
> case MEMORY_DEVICE_PCI_P2PDMA:
>
No strong opinion (@Dan?), I do wonder if a separate type would make sense.
--
Thanks,
David / dhildenb
[PATCH v4 1/2] memremap: rename MEMORY_DEVICE_DEVDAX to MEMORY_DEVICE_GENERIC
This is in preparation for the logic behind MEMORY_DEVICE_DEVDAX also
being used by non DAX devices.
No functional change intended.
Signed-off-by: Roger Pau Monné
---
Cc: Dan Williams
Cc: Vishal Verma
Cc: Dave Jiang
Cc: Andrew Morton
Cc: Jason Gunthorpe
Cc: Ira Weiny
Cc: "Aneesh Kumar K.V"
Cc: Johannes Thumshirn
Cc: Logan Gunthorpe
Cc: linux-nvd...@lists.01.org
Cc: xen-de...@lists.xenproject.org
Cc: linux...@kvack.org
---
drivers/dax/device.c | 2 +-
include/linux/memremap.h | 9 -
mm/memremap.c| 2 +-
3 files changed, 6 insertions(+), 7 deletions(-)
diff --git a/drivers/dax/device.c b/drivers/dax/device.c
index 4c0af2eb7e19..1e89513f3c59 100644
--- a/drivers/dax/device.c
+++ b/drivers/dax/device.c
@@ -429,7 +429,7 @@ int dev_dax_probe(struct device *dev)
return -EBUSY;
}
- dev_dax->pgmap.type = MEMORY_DEVICE_DEVDAX;
+ dev_dax->pgmap.type = MEMORY_DEVICE_GENERIC;
addr = devm_memremap_pages(dev, _dax->pgmap);
if (IS_ERR(addr))
return PTR_ERR(addr);
diff --git a/include/linux/memremap.h b/include/linux/memremap.h
index 5f5b2df06e61..e5862746751b 100644
--- a/include/linux/memremap.h
+++ b/include/linux/memremap.h
@@ -46,11 +46,10 @@ struct vmem_altmap {
* wakeup is used to coordinate physical address space management (ex:
* fs truncate/hole punch) vs pinned pages (ex: device dma).
*
- * MEMORY_DEVICE_DEVDAX:
+ * MEMORY_DEVICE_GENERIC:
* Host memory that has similar access semantics as System RAM i.e. DMA
- * coherent and supports page pinning. In contrast to
- * MEMORY_DEVICE_FS_DAX, this memory is access via a device-dax
- * character device.
+ * coherent and supports page pinning. This is for example used by DAX devices
+ * that expose memory using a character device.
*
* MEMORY_DEVICE_PCI_P2PDMA:
* Device memory residing in a PCI BAR intended for use with Peer-to-Peer
@@ -60,7 +59,7 @@ enum memory_type {
/* 0 is reserved to catch uninitialized type fields */
MEMORY_DEVICE_PRIVATE = 1,
MEMORY_DEVICE_FS_DAX,
- MEMORY_DEVICE_DEVDAX,
+ MEMORY_DEVICE_GENERIC,
MEMORY_DEVICE_PCI_P2PDMA,
};
diff --git a/mm/memremap.c b/mm/memremap.c
index 03e38b7a38f1..006dace60b1a 100644
--- a/mm/memremap.c
+++ b/mm/memremap.c
@@ -216,7 +216,7 @@ void *memremap_pages(struct dev_pagemap *pgmap, int nid)
return ERR_PTR(-EINVAL);
}
break;
- case MEMORY_DEVICE_DEVDAX:
+ case MEMORY_DEVICE_GENERIC:
need_devmap_managed = false;
break;
case MEMORY_DEVICE_PCI_P2PDMA:
--
2.28.0
