On Tue, Nov 19, 2019 at 7:19 PM Aneesh Kumar K.V
wrote:
>
> On 11/19/19 11:28 PM, Dan Williams wrote:
> > On Mon, Nov 18, 2019 at 1:52 AM Aneesh Kumar K.V
> > wrote:
> >>
> >> Dan Williams writes:
> >>
> >>> On Sat, Nov 16, 2019 at 4:15 AM Aneesh Kumar K.V
> >>> wrote:
>
> >>
> >>
> >
On 11/19/19 11:28 PM, Dan Williams wrote:
On Mon, Nov 18, 2019 at 1:52 AM Aneesh Kumar K.V
wrote:
Dan Williams writes:
On Sat, Nov 16, 2019 at 4:15 AM Aneesh Kumar K.V
wrote:
Considering the direct-map map size is not going to be user selectable,
do you agree that we can skip
On Mon, Nov 18, 2019 at 1:52 AM Aneesh Kumar K.V
wrote:
>
> Dan Williams writes:
>
> > On Sat, Nov 16, 2019 at 4:15 AM Aneesh Kumar K.V
> > wrote:
> >>
>
>
>
>
> >>
> >> Considering the direct-map map size is not going to be user selectable,
> >> do you agree that we can skip the above step
Dan Williams writes:
> On Sat, Nov 16, 2019 at 4:15 AM Aneesh Kumar K.V
> wrote:
>>
>>
>> Considering the direct-map map size is not going to be user selectable,
>> do you agree that we can skip the above step 0 configuration you
>> suggested.
>>
>> The changes proposed in the patch seri
On Sat, Nov 16, 2019 at 4:15 AM Aneesh Kumar K.V
wrote:
>
>
> Hi Dan,
>
> "Aneesh Kumar K.V" writes:
>
> > Dan Williams writes:
> >
> >> On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
> >> wrote:
> >> [..]
> >>> > True, for the pfn device and the device-dax mapping size, but I'm
> >>> > sugg
Hi Dan,
"Aneesh Kumar K.V" writes:
> Dan Williams writes:
>
>> On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
>> wrote:
>> [..]
>>> > True, for the pfn device and the device-dax mapping size, but I'm
>>> > suggesting adding another instance of alignment control at the raw
>>> > namespace
Dan Williams writes:
> On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
> wrote:
> [..]
>> > True, for the pfn device and the device-dax mapping size, but I'm
>> > suggesting adding another instance of alignment control at the raw
>> > namespace level. That would need to be disconnected from th
On Thu, Oct 31, 2019 at 1:38 AM Aneesh Kumar K.V
wrote:
>
> On 10/31/19 12:00 PM, Dan Williams wrote:
> > On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
> > wrote:
> > [..]
>
> >
> >
> > All that said, the x86 vmemmap_populate() falls back to use small
> > pages in some case to get around this
On 10/31/19 12:00 PM, Dan Williams wrote:
On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
wrote:
[..]
All that said, the x86 vmemmap_populate() falls back to use small
pages in some case to get around this constraint. Can't powerpc do the
same? It would seem to be less work than the above
On Wed, Oct 30, 2019 at 10:35 PM Aneesh Kumar K.V
wrote:
[..]
> > True, for the pfn device and the device-dax mapping size, but I'm
> > suggesting adding another instance of alignment control at the raw
> > namespace level. That would need to be disconnected from the
> > device-dax page mapping gr
On 10/29/19 11:00 AM, Dan Williams wrote:
On Mon, Oct 28, 2019 at 9:35 PM Aneesh Kumar K.V
wrote:
On 10/29/19 4:38 AM, Dan Williams wrote:
On Mon, Oct 28, 2019 at 2:48 AM Aneesh Kumar K.V
wrote:
The page size used to map the namespace is arch dependent. For example
architectures like ppc64
On Mon, Oct 28, 2019 at 9:35 PM Aneesh Kumar K.V
wrote:
>
> On 10/29/19 4:38 AM, Dan Williams wrote:
> > On Mon, Oct 28, 2019 at 2:48 AM Aneesh Kumar K.V
> > wrote:
> >>
> >> The page size used to map the namespace is arch dependent. For example
> >> architectures like ppc64 use 16MB page size fo
On 10/29/19 2:51 AM, Ira Weiny wrote:
diff --git a/arch/powerpc/lib/pmem.c b/arch/powerpc/lib/pmem.c
index 377712e85605..2e661a08dae5 100644
--- a/arch/powerpc/lib/pmem.c
+++ b/arch/powerpc/lib/pmem.c
@@ -17,14 +17,31 @@ void arch_wb_cache_pmem(void *addr, size_t size)
unsigned long start
On 10/29/19 4:38 AM, Dan Williams wrote:
On Mon, Oct 28, 2019 at 2:48 AM Aneesh Kumar K.V
wrote:
The page size used to map the namespace is arch dependent. For example
architectures like ppc64 use 16MB page size for direct-mapping. If the namespace
size is not aligned to the mapping page size,
On Mon, Oct 28, 2019 at 2:48 AM Aneesh Kumar K.V
wrote:
>
> The page size used to map the namespace is arch dependent. For example
> architectures like ppc64 use 16MB page size for direct-mapping. If the
> namespace
> size is not aligned to the mapping page size, we can observe kernel crash
> dur
> diff --git a/arch/powerpc/lib/pmem.c b/arch/powerpc/lib/pmem.c
> index 377712e85605..2e661a08dae5 100644
> --- a/arch/powerpc/lib/pmem.c
> +++ b/arch/powerpc/lib/pmem.c
> @@ -17,14 +17,31 @@ void arch_wb_cache_pmem(void *addr, size_t size)
> unsigned long start = (unsigned long) addr;
>
The page size used to map the namespace is arch dependent. For example
architectures like ppc64 use 16MB page size for direct-mapping. If the namespace
size is not aligned to the mapping page size, we can observe kernel crash
during namespace init and destroy.
This is due to kernel doing partial m
17 matches
Mail list logo
