net.ipv4.tcp_rmem does override net.core.rmem_max?

net.ipv4.tcp_rmem does override net.core.rmem_max?

[Niklas Hambüchen]

`man tcp` says `net.ipv4.tcp_rmem`'s max value *does not override*
`net.core.rmem_max` (see [1] [2]) but I found in practice it does.

The person in this thread
http://stackoverflow.com/questions/31546835/tcp-receiving-window-size-higher-than-net-core-rmem-max/35438236
confirms my finding.

Is the man page (and ip-sysctl.txt) outdated?

Thanks for clarifying!

Niklas

[1]
https://github.com/mkerrisk/man-pages/blob/af54e3b7163998cbe1f487f4dca040615fbf949b/man7/tcp.7#L745
[2]
https://github.com/torvalds/linux/blob/34229b277480f46c1e9a19f027f30b074512e68b/Documentation/networking/ip-sysctl.txt#L512
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: net.ipv4.tcp_rmem does override net.core.rmem_max?

[Niklas Hambüchen]

An answer to the question I just posted,

  http://stackoverflow.com/a/35438236/263061

suggests that neither is the case, but in fact the kernel takes the
maximum of the two values.
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: net.ipv4.tcp_rmem does override net.core.rmem_max?

[Niklas Hambüchen]

Or, there is the following interpretation:

  "This  value  does  not  override   the   global
  net.core.rmem_max."

might mean

  "This value does not override the global
  net.core.rmem_max to the extent that it does
  not define the upper users can set with SO_RCVBUF.
  It does, however, override net.core.rmem_max if
  SO_RCVBUF is not used."

This would be consistent with my findings.

I'm not 100% confident about whether this is what the code does though,
see my comment
http://stackoverflow.com/questions/31546835/tcp-receiving-window-size-higher-than-net-core-rmem-max/35438236#comment58576526_35438236
for details on that.

On 16/02/16 18:05, Niklas Hambüchen wrote:
> An answer to the question I just posted,
> 
>   http://stackoverflow.com/a/35438236/263061
> 
> suggests that neither is the case, but in fact the kernel takes the
> maximum of the two values.
> 
--
To unsubscribe from this list: send the line "unsubscribe linux-man" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html