Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-04 Thread Dongli Zhang
On 11/4/20 4:51 AM, Eric Dumazet wrote: > > > On 11/4/20 1:36 PM, Matthew Wilcox wrote: >> On Wed, Nov 04, 2020 at 09:50:30AM +0100, Eric Dumazet wrote: >>> On 11/4/20 2:16 AM, Rama Nichanamatlu wrote: > Thanks for providing the numbers.  Do you think that dropping (up to) > 7 packets

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-04 Thread Eric Dumazet
On 11/4/20 1:36 PM, Matthew Wilcox wrote: > On Wed, Nov 04, 2020 at 09:50:30AM +0100, Eric Dumazet wrote: >> On 11/4/20 2:16 AM, Rama Nichanamatlu wrote: Thanks for providing the numbers.  Do you think that dropping (up to) 7 packets is acceptable? >>> >>> net.ipv4.tcp_syn_retries = 6

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-04 Thread Matthew Wilcox
On Wed, Nov 04, 2020 at 09:50:30AM +0100, Eric Dumazet wrote: > On 11/4/20 2:16 AM, Rama Nichanamatlu wrote: > >> Thanks for providing the numbers.  Do you think that dropping (up to) > >> 7 packets is acceptable? > > > > net.ipv4.tcp_syn_retries = 6 > > > > tcp clients wouldn't even get that far

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-04 Thread Eric Dumazet
On 11/4/20 2:16 AM, Rama Nichanamatlu wrote: >> Thanks for providing the numbers.  Do you think that dropping (up to) >> 7 packets is acceptable? > > net.ipv4.tcp_syn_retries = 6 > > tcp clients wouldn't even get that far leading to connect establish issues. This does not really matter. If ho

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Rama Nichanamatlu
Thanks for providing the numbers. Do you think that dropping (up to) 7 packets is acceptable? net.ipv4.tcp_syn_retries = 6 tcp clients wouldn't even get that far leading to connect establish issues. -rama On Tue, Nov 03, 2020 at 09:15:41PM +, Matthew Wilcox wrote: On Tue, Nov 03, 2020 at

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Dongli Zhang
On 11/3/20 1:15 PM, Matthew Wilcox wrote: > On Tue, Nov 03, 2020 at 12:57:33PM -0800, Dongli Zhang wrote: >> On 11/3/20 12:35 PM, Matthew Wilcox wrote: >>> On Tue, Nov 03, 2020 at 11:32:39AM -0800, Dongli Zhang wrote: However, once kernel is not under memory pressure any longer (suppose lar

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Dongli Zhang
Hi Matthew, On 11/3/20 12:35 PM, Matthew Wilcox wrote: > On Tue, Nov 03, 2020 at 11:32:39AM -0800, Dongli Zhang wrote: >> The ethernet driver may allocates skb (and skb->data) via napi_alloc_skb(). >> This ends up to page_frag_alloc() to allocate skb->data from >> page_frag_cache->va. >> >> During

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Matthew Wilcox
On Tue, Nov 03, 2020 at 12:57:33PM -0800, Dongli Zhang wrote: > On 11/3/20 12:35 PM, Matthew Wilcox wrote: > > On Tue, Nov 03, 2020 at 11:32:39AM -0800, Dongli Zhang wrote: > >> However, once kernel is not under memory pressure any longer (suppose large > >> amount of memory pages are just reclaime

Re: [PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Matthew Wilcox
On Tue, Nov 03, 2020 at 11:32:39AM -0800, Dongli Zhang wrote: > The ethernet driver may allocates skb (and skb->data) via napi_alloc_skb(). > This ends up to page_frag_alloc() to allocate skb->data from > page_frag_cache->va. > > During the memory pressure, page_frag_cache->va may be allocated as

[PATCH 1/1] mm: avoid re-using pfmemalloc page in page_frag_alloc()

2020-11-03 Thread Dongli Zhang
The ethernet driver may allocates skb (and skb->data) via napi_alloc_skb(). This ends up to page_frag_alloc() to allocate skb->data from page_frag_cache->va. During the memory pressure, page_frag_cache->va may be allocated as pfmemalloc page. As a result, the skb->pfmemalloc is always true as skb-