<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux.git/net/tls, branch v5.4-rc2</title>
<subtitle>Linux kernel source tree</subtitle>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/'/>
<entry>
<title>net/tls: align non temporal copy to cache lines</title>
<updated>2019-09-07T16:10:34+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-07T05:30:00+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=e681cc603a79b5907a7388e1d4587779b0fe8ff2'/>
<id>e681cc603a79b5907a7388e1d4587779b0fe8ff2</id>
<content type='text'>
Unlike normal TCP code TLS has to touch the cache lines
it copies into to fill header info. On memory-heavy workloads
having non temporal stores and normal accesses targeting
the same cache line leads to significant overhead.

Measured 3% overhead running 3600 round robin connections
with additional memory heavy workload.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Unlike normal TCP code TLS has to touch the cache lines
it copies into to fill header info. On memory-heavy workloads
having non temporal stores and normal accesses targeting
the same cache line leads to significant overhead.

Measured 3% overhead running 3600 round robin connections
with additional memory heavy workload.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: remove the record tail optimization</title>
<updated>2019-09-07T16:10:34+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-07T05:29:59+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=e7b159a48ba6f7c243881c7ef3afa6e8785c0826'/>
<id>e7b159a48ba6f7c243881c7ef3afa6e8785c0826</id>
<content type='text'>
For TLS device offload the tag/message authentication code are
filled in by the device. The kernel merely reserves space for
them. Because device overwrites it, the contents of the tag make
do no matter. Current code tries to save space by reusing the
header as the tag. This, however, leads to an additional frag
being created and defeats buffer coalescing (which trickles
all the way down to the drivers).

Remove this optimization, and try to allocate the space for
the tag in the usual way, leave the memory uninitialized.
If memory allocation fails rewind the record pointer so that
we use the already copied user data as tag.

Note that the optimization was actually buggy, as the tag
for TLS 1.2 is 16 bytes, but header is just 13, so the reuse
may had looked past the end of the page..

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
For TLS device offload the tag/message authentication code are
filled in by the device. The kernel merely reserves space for
them. Because device overwrites it, the contents of the tag make
do no matter. Current code tries to save space by reusing the
header as the tag. This, however, leads to an additional frag
being created and defeats buffer coalescing (which trickles
all the way down to the drivers).

Remove this optimization, and try to allocate the space for
the tag in the usual way, leave the memory uninitialized.
If memory allocation fails rewind the record pointer so that
we use the already copied user data as tag.

Note that the optimization was actually buggy, as the tag
for TLS 1.2 is 16 bytes, but header is just 13, so the reuse
may had looked past the end of the page..

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: use RCU for the adder to the offload record list</title>
<updated>2019-09-07T16:10:34+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-07T05:29:58+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=d4774ac0d49ae92c5176c9848db555e89a5a4e45'/>
<id>d4774ac0d49ae92c5176c9848db555e89a5a4e45</id>
<content type='text'>
All modifications to TLS record list happen under the socket
lock. Since records form an ordered queue readers are only
concerned about elements being removed, additions can happen
concurrently.

Use RCU primitives to ensure the correct access types
(READ_ONCE/WRITE_ONCE).

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
All modifications to TLS record list happen under the socket
lock. Since records form an ordered queue readers are only
concerned about elements being removed, additions can happen
concurrently.

Use RCU primitives to ensure the correct access types
(READ_ONCE/WRITE_ONCE).

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: unref frags in order</title>
<updated>2019-09-07T16:10:34+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-07T05:29:57+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=7ccd451912dd6407957676ae584248bdadc97b38'/>
<id>7ccd451912dd6407957676ae584248bdadc97b38</id>
<content type='text'>
It's generally more cache friendly to walk arrays in order,
especially those which are likely not in cache.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
It's generally more cache friendly to walk arrays in order,
especially those which are likely not in cache.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: dedup the record cleanup</title>
<updated>2019-09-05T07:49:49+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-03T04:31:06+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=6e3d02b670eafce69aa64457b87c3b17ff216d23'/>
<id>6e3d02b670eafce69aa64457b87c3b17ff216d23</id>
<content type='text'>
If retransmit record hint fall into the cleanup window we will
free it by just walking the list. No need to duplicate the code.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
If retransmit record hint fall into the cleanup window we will
free it by just walking the list. No need to duplicate the code.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: clean up the number of #ifdefs for CONFIG_TLS_DEVICE</title>
<updated>2019-09-05T07:49:49+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-03T04:31:05+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=be2fbc155fc8c0ff6e499753354d965cd9cf1bb0'/>
<id>be2fbc155fc8c0ff6e499753354d965cd9cf1bb0</id>
<content type='text'>
TLS code has a number of #ifdefs which make the code a little
harder to follow. Recent fixes removed the ifdef around the
TLS_HW define, so we can switch to the often used pattern
of defining tls_device functions as empty static inlines
in the header when CONFIG_TLS_DEVICE=n.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Acked-by: John Fastabend &lt;john.fastabend@gmail.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
TLS code has a number of #ifdefs which make the code a little
harder to follow. Recent fixes removed the ifdef around the
TLS_HW define, so we can switch to the often used pattern
of defining tls_device functions as empty static inlines
in the header when CONFIG_TLS_DEVICE=n.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Acked-by: John Fastabend &lt;john.fastabend@gmail.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: narrow down the critical area of device_offload_lock</title>
<updated>2019-09-05T07:49:49+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-03T04:31:04+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=3544c98acd09b3b40e86f015f7df75a7d2d72a5c'/>
<id>3544c98acd09b3b40e86f015f7df75a7d2d72a5c</id>
<content type='text'>
On setsockopt path we need to hold device_offload_lock from
the moment we check netdev is up until the context is fully
ready to be added to the tls_device_list.

No need to hold it around the get_netdev_for_sock().
Change the code and remove the confusing comment.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
On setsockopt path we need to hold device_offload_lock from
the moment we check netdev is up until the context is fully
ready to be added to the tls_device_list.

No need to hold it around the get_netdev_for_sock().
Change the code and remove the confusing comment.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: don't jump to return</title>
<updated>2019-09-05T07:49:49+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-03T04:31:03+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=90962b4894f560432e5538e9a354ee708f8beb3d'/>
<id>90962b4894f560432e5538e9a354ee708f8beb3d</id>
<content type='text'>
Reusing parts of error path for normal exit will make
next commit harder to read, untangle the two.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Reusing parts of error path for normal exit will make
next commit harder to read, untangle the two.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net/tls: use the full sk_proto pointer</title>
<updated>2019-09-05T07:49:49+00:00</updated>
<author>
<name>Jakub Kicinski</name>
<email>jakub.kicinski@netronome.com</email>
</author>
<published>2019-09-03T04:31:02+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=be7bbea114d6ab2688b9e59cd24a306d21e51c27'/>
<id>be7bbea114d6ab2688b9e59cd24a306d21e51c27</id>
<content type='text'>
Since we already have the pointer to the full original sk_proto
stored use that instead of storing all individual callback
pointers as well.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Acked-by: John Fastabend &lt;john.fastabend@gmail.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Since we already have the pointer to the full original sk_proto
stored use that instead of storing all individual callback
pointers as well.

Signed-off-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Reviewed-by: John Hurley &lt;john.hurley@netronome.com&gt;
Reviewed-by: Dirk van der Merwe &lt;dirk.vandermerwe@netronome.com&gt;
Acked-by: John Fastabend &lt;john.fastabend@gmail.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>net: tls: export protocol version, cipher, tx_conf/rx_conf to socket diag</title>
<updated>2019-09-01T06:44:28+00:00</updated>
<author>
<name>Davide Caratti</name>
<email>dcaratti@redhat.com</email>
</author>
<published>2019-08-30T10:25:49+00:00</published>
<link rel='alternate' type='text/html' href='https://git.tavy.me/linux.git/commit/?id=26811cc9f55acf835f7fdadc5ff2bbd6f06bc3ac'/>
<id>26811cc9f55acf835f7fdadc5ff2bbd6f06bc3ac</id>
<content type='text'>
When an application configures kernel TLS on top of a TCP socket, it's
now possible for inet_diag_handler() to collect information regarding the
protocol version, the cipher type and TX / RX configuration, in case
INET_DIAG_INFO is requested.

Signed-off-by: Davide Caratti &lt;dcaratti@redhat.com&gt;
Reviewed-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
When an application configures kernel TLS on top of a TCP socket, it's
now possible for inet_diag_handler() to collect information regarding the
protocol version, the cipher type and TX / RX configuration, in case
INET_DIAG_INFO is requested.

Signed-off-by: Davide Caratti &lt;dcaratti@redhat.com&gt;
Reviewed-by: Jakub Kicinski &lt;jakub.kicinski@netronome.com&gt;
Signed-off-by: David S. Miller &lt;davem@davemloft.net&gt;
</pre>
</div>
</content>
</entry>
</feed>
