Re: sock_alloc_send_skb(): wrong GFP flags?

netdev

Re: sock_alloc_send_skb(): wrong GFP flags?

To:	manfreds@xxxxxxxxxxxxxxxx (Manfred Spraul)
Subject:	Re: sock_alloc_send_skb(): wrong GFP flags?
From:	Alan Cox <alan@xxxxxxxxxxxxxxxxxxx>
Date:	Wed, 24 May 2000 22:18:38 +0100 (BST)
Cc:	davem@xxxxxxxxxx, linux-net@xxxxxxxxxxxxxxxx, netdev@xxxxxxxxxxx
In-reply-to:	<392C38BF.F45193DA@colorfullife.com> from "Manfred Spraul" at May 24, 2000 10:17:03 PM
Sender:	owner-netdev@xxxxxxxxxxx

> IMHO the flags are wrong:
> Someone modified GFP_BUFFER between 2.2.15 and 2.3.99-pre9:
> 
> was:
>       GFP_BUFFER == GFP_MED|GFP_WAIT
> now:
>       GFP_BUFFER == GFP_HIGH|GFP_WAIT
> 
> It will eat the memory for atomic allocations, and it will sleep instead
> of downgrading to fallback.

It needs to allocate non atomic memory only 

> Should I write a patch?

Please. I had completely missed that

> I'm sure that the actual data kmalloc fails far more often than the
> skb_head alloc. We should allocate the data area first, and then the
> skb_head.

True

[More with this subject...]

<Prev in Thread]	Current Thread	[Next in Thread>
sock_alloc_send_skb(): wrong GFP flags?, Manfred Spraul Re: sock_alloc_send_skb(): wrong GFP flags?, Alan Cox <=

Previous by Date:	sock_alloc_send_skb(): wrong GFP flags?, Manfred Spraul
Next by Date:	Re: tx_timeout and timer serialisation, kuznet
Previous by Thread:	sock_alloc_send_skb(): wrong GFP flags?, Manfred Spraul
Next by Thread:	tap devices, ddp, multicast, Olaf Hering
Indexes:	[Date] [Thread] [Top] [All Lists]