linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] mm/cma: get nid from physical address
@ 2024-05-16  9:17 Yajun Deng
  2024-05-16 10:03 ` Kefeng Wang
  0 siblings, 1 reply; 3+ messages in thread
From: Yajun Deng @ 2024-05-16  9:17 UTC (permalink / raw)
  To: akpm; +Cc: linux-mm, linux-kernel, Yajun Deng

The nid passed to cma_declare_contiguous_nid() may be NUMA_NO_NODE,
which is not the actual nid. To get the correct nid, we can get the nid
from physical address.

Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
---
 mm/cma.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/mm/cma.c b/mm/cma.c
index 3e9724716bad..be6cdde32944 100644
--- a/mm/cma.c
+++ b/mm/cma.c
@@ -361,6 +361,7 @@ int __init cma_declare_contiguous_nid(phys_addr_t base,
 		kmemleak_ignore_phys(addr);
 		base = addr;
 	}
+	nid = early_pfn_to_nid(PHYS_PFN(base));
 
 	ret = cma_init_reserved_mem(base, size, order_per_bit, name, res_cma);
 	if (ret)
-- 
2.25.1



^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] mm/cma: get nid from physical address
  2024-05-16  9:17 [PATCH] mm/cma: get nid from physical address Yajun Deng
@ 2024-05-16 10:03 ` Kefeng Wang
  2024-05-16 11:16   ` Yajun Deng
  0 siblings, 1 reply; 3+ messages in thread
From: Kefeng Wang @ 2024-05-16 10:03 UTC (permalink / raw)
  To: Yajun Deng, akpm; +Cc: linux-mm, linux-kernel



On 2024/5/16 17:17, Yajun Deng wrote:
> The nid passed to cma_declare_contiguous_nid() may be NUMA_NO_NODE,
> which is not the actual nid. To get the correct nid, we can get the nid
> from physical address.
> 

Please check

https://lore.kernel.org/linux-riscv/47437c2b-5946-41c6-ad1b-cc03329eb230@huawei.com/

> Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
> ---
>   mm/cma.c | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/mm/cma.c b/mm/cma.c
> index 3e9724716bad..be6cdde32944 100644
> --- a/mm/cma.c
> +++ b/mm/cma.c
> @@ -361,6 +361,7 @@ int __init cma_declare_contiguous_nid(phys_addr_t base,
>   		kmemleak_ignore_phys(addr);
>   		base = addr;
>   	}
> +	nid = early_pfn_to_nid(PHYS_PFN(base));
>   
>   	ret = cma_init_reserved_mem(base, size, order_per_bit, name, res_cma);
>   	if (ret)


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] mm/cma: get nid from physical address
  2024-05-16 10:03 ` Kefeng Wang
@ 2024-05-16 11:16   ` Yajun Deng
  0 siblings, 0 replies; 3+ messages in thread
From: Yajun Deng @ 2024-05-16 11:16 UTC (permalink / raw)
  To: Kefeng Wang, akpm; +Cc: linux-mm, linux-kernel

May 16, 2024 at 6:03 PM, "Kefeng Wang" <wangkefeng.wang@huawei.com> wrote:



> 
> On 2024/5/16 17:17, Yajun Deng wrote:
> 
> > 
> > The nid passed to cma_declare_contiguous_nid() may be NUMA_NO_NODE,
> > 
> >  which is not the actual nid. To get the correct nid, we can get the nid
> > 
> >  from physical address.
> > 
> 
> Please check
> 
> https://lore.kernel.org/linux-riscv/47437c2b-5946-41c6-ad1b-cc03329eb230@huawei.com/
> 

Okay, thanks!
> > 
> > Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
> > 
> >  ---
> > 
> >  mm/cma.c | 1 +
> > 
> >  1 file changed, 1 insertion(+)
> > 
> >  diff --git a/mm/cma.c b/mm/cma.c
> > 
> >  index 3e9724716bad..be6cdde32944 100644
> > 
> >  --- a/mm/cma.c
> > 
> >  +++ b/mm/cma.c
> > 
> >  @@ -361,6 +361,7 @@ int __init cma_declare_contiguous_nid(phys_addr_t base,
> > 
> >  kmemleak_ignore_phys(addr);
> > 
> >  base = addr;
> > 
> >  }
> > 
> >  + nid = early_pfn_to_nid(PHYS_PFN(base));
> > 
> >  > ret = cma_init_reserved_mem(base, size, order_per_bit, name, res_cma);
> > 
> >  if (ret)
> >
>


^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2024-05-16 11:16 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-05-16  9:17 [PATCH] mm/cma: get nid from physical address Yajun Deng
2024-05-16 10:03 ` Kefeng Wang
2024-05-16 11:16   ` Yajun Deng

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).