[PATCH 4.15-rc8] net/core: Increase default optmem_max limit

[PATCH 4.15-rc8] net/core: Increase default optmem_max limit

[Björn 'besser82' Esser]

[-- Attachment #1: Type: text/plain, Size: 1435 bytes --]

With the new Linux Kernel Crypto API User Space Interface and
its underlying socket interface, the current default value for
`net.core.optmem_max` can be exhausted pretty quick.
On 32 bit systems it is not even enough for sending 16 IOVECs
at once to the socket interface.

To provide consumers of this new user space interface a
sufficient and reasonable maximum ancillary buffer size per
socket by default, the limit is increased to four times of the
previous setting:

  * 32 bit systems:  from 10240 bytes to 40960 bytes
  * 64 bit systems:  from 20480 bytes to 81920 bytes

This allows for sending 32/64 (32/64 bit) parallel IOVECs at
once to the socket interface, which should be enough for use
in real world applications.

Signed-off-by: Björn Esser <besser82@fedoraproject.org>
---

Index: linux-4.15/net/core/sock.c
===================================================================
--- linux-4.15.orig/net/core/sock.c
+++ linux-4.15/net/core/sock.c
@@ -316,7 +316,7 @@ __u32 sysctl_wmem_default __read_mostly
 __u32 sysctl_rmem_default __read_mostly = SK_RMEM_MAX;
 
 /* Maximal space eaten by iovec or
ancillary data plus some space */
-int sysctl_optmem_max __read_mostly = sizeof(unsigned long)*(2*UIO_MAXIOV+512);
+int sysctl_optmem_max __read_mostly
= sizeof(unsigned long)*4*(2*UIO_MAXIOV+512);
 EXPORT_SYMBOL(sysctl_optmem_max);
 
 int sysctl_tstamp_allow_data __read_mostly = 1;

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

Re: [PATCH 4.15-rc8] net/core: Increase default optmem_max limit

[David Miller]

From: Björn 'besser82' Esser <besser82@fedoraproject.org>
Date: Tue, 16 Jan 2018 00:59:11 +0100

> @@ -316,7 +316,7 @@ __u32 sysctl_wmem_default __read_mostly
>  __u32 sysctl_rmem_default __read_mostly = SK_RMEM_MAX;
>  
>  /* Maximal
> space eaten by iovec or ancillary data plus some space */


This patch is corrupted by your email client, it has chopped up long
lines and turned TABs into a series of SPACE characters.

> -int sysctl_optmem_max __read_mostly = sizeof(unsigned long)*(2*UIO_MAXIOV+512);
> +int
> sysctl_optmem_max __read_mostly = sizeof(unsigned long)*4*(2*UIO_MAXIOV+512);
>  EXPORT_SYMBOL(sysctl_optmem_max);

Please style this in a more appropriate manner.

Thank you.

Re: [PATCH 4.15-rc8] net/core: Increase default optmem_max limit

[Björn 'besser82' Esser]

With the new Linux Kernel Crypto API User Space Interface and
its underlying socket interface, the current default value for
`net.core.optmem_max` can be exhausted pretty quick.
On 32 bit systems it is not even enough for sending 16 IOVECs
at once to the socket interface.

To provide consumers of this new user space interface a
sufficient and reasonable maximum ancillary buffer size per
socket by default, the limit is increased to four times of the
previous setting:

  * 32 bit systems:  from 10240 bytes to 40960 bytes
  * 64 bit systems:  from 20480 bytes to 81920 bytes

This allows for sending 32/64 (32/64 bit) parallel IOVECs at
once to the socket interface, which should be enough for use
in real world applications.

Signed-off-by: Björn Esser <besser82@fedoraproject.org>
---

Index: linux-4.15/net/core/sock.c
===================================================================
--- linux-4.15.orig/net/core/sock.c
+++ linux-4.15/net/core/sock.c
@@ -316,7 +316,7 @@ __u32 sysctl_wmem_default __read_mostly
 __u32 sysctl_rmem_default __read_mostly = SK_RMEM_MAX;
 
 /* Maximal space eaten by iovec or ancillary data plus some space */
-int sysctl_optmem_max __read_mostly = sizeof(unsigned long)*(2*UIO_MAXIOV+512);
+int sysctl_optmem_max __read_mostly = sizeof(unsigned long)*4*(2*UIO_MAXIOV+512);
 EXPORT_SYMBOL(sysctl_optmem_max);
 
 int sysctl_tstamp_allow_data __read_mostly = 1;

[PATCH 4.15-rc8] net/core: Increase default optmem_max limit

[Björn 'besser82' Esser]

With the new Linux Kernel Crypto API User Space Interface and
its underlying socket interface, the current default value for
`net.core.optmem_max` can be exhausted pretty quick.
On 32 bit systems it is not even enough for sending 16 IOVECs
at once to the socket interface.

To provide consumers of this new user space interface a
sufficient and reasonable maximum ancillary buffer size per
socket by default, the limit is increased to four times of the
previous setting:

  * 32 bit systems:  from 10240 bytes to 40960 bytes
  * 64 bit systems:  from 20480 bytes to 81920 bytes

This allows for sending 32/64 (32/64 bit) parallel IOVECs at
once to the socket interface, which should be enough for use
in real world applications.

Signed-off-by: Björn Esser <besser82@fedoraproject.org>
---

Index: linux-4.15/net/core/sock.c
===================================================================
--- linux-4.15.orig/net/core/sock.c
+++ linux-
4.15/net/core/sock.c
@@ -316,7 +316,7 @@ __u32 sysctl_wmem_default __read_mostly
 __u32 sysctl_rmem_default __read_mostly = SK_RMEM_MAX;
 
 /* Maximal
space eaten by iovec or ancillary data plus some space */
-int sysctl_optmem_max __read_mostly = sizeof(unsigned long)*(2*UIO_MAXIOV+512);
+int
sysctl_optmem_max __read_mostly = sizeof(unsigned long)*4*(2*UIO_MAXIOV+512);
 EXPORT_SYMBOL(sysctl_optmem_max);
 
 int sysctl_tstamp_allow_data
__read_mostly = 1;