SteamSocket::eraseFirstInputBytes() removes from the beginning
of std::vector, which is generally slow. If the buffer becomes
too big, which it may under a load, then the function will get
slow, which in turn will likely lead to the buffer getting even
bigger because of accumulated backlog.
The Buffer class is optimized for removal at the beginning,
so use it instead of std::vector, including some API additions
for it to be an in-place replacement where it's used.
Signed-off-by: Luboš Luňák <l.lunak@collabora.com>
Change-Id: I4cf7ec56c908c7d3df391dc3f8e230ad32abb162
test/test.cpp and wsd/Storage.cpp has the same pattern.
Signed-off-by: Miklos Vajna <vmiklos@collabora.com>
Change-Id: Ic1dfbd3a5368eb9831bde9d77ea1bb813834d42b
There is one EndSession record for each session that edited the same
document. That is not an error.
Signed-off-by: Tor Lillqvist <tml@collabora.com>
Change-Id: I9042411ad465fe07a8f88a29923f0dc051c755aa
Signed-off-by: Tor Lillqvist <tml@collabora.com>
I otherwise get values like this:
epochFile=18682409
deltaCurrent=34
rec.getTimestampNs()=18682409
deltaFile=0
delay=4294967262
Leading to the idiotic:
Sleeping for 4294967 ms
Signed-off-by: Tor Lillqvist <tml@collabora.com>
Change-Id: I06530c31bc70bb7dcf3cd34c4684aa9f1f8036cb
Move the generic dummy implementation of
TraceEvent::emitOneRecording() to a source file of its own. (That is
the one which is used in test and tool executables.)
Signed-off-by: Tor Lillqvist <tml@collabora.com>
Change-Id: I81cab07e5a6852b42d278a5446c13c3825cf546e
Modelled on how it is done in core. ProfileZone is derived from
NamedEvent which is derived from TraceEvent. Here we don't keep any
separate ProfileZone.hpp, though.
This was needed to introduce generation of "instant" events here, too.
Signed-off-by: Tor Lillqvist <tml@collabora.com>
Change-Id: I6583134e96001641c50339deb4197fca6ab7d5d5
Don't bother collecting them into a vector, like in core. Instead,
just call a static member function ProfileZone::addOneRecording() for
each Event Trace object (string) to be emitted. That function then
needs to be implemented separately in each executable. In WSD it logs
the object to the Event Trace file. In Kit it sends it to WSD for
logging. In Unit tests we use a dummy implementation.
(If Event Trace logging is not enabled at all in loolwsd.xml (the
default), nothing is done, of course.)
When receiving the "traceeventrecording start" or "traceeventrecording
stop" message from the client, turn ProfileZone recoring on or off in
the WSD and Kit process. (Probably in WSD the flag should be
per-client.)
Change-Id: Ie1127d65dd44ed77e7eeab4b0f0a90cce95dc4a2
Signed-off-by: Tor Lillqvist <tml@collabora.com>
When writing to the socket, it's always more efficient
to fill the buffer up to the hardware limit for each
write. This is doubly important for efficiency with
SSL, due to the overhead of encrypting multiple
small buffers instead of one large one.
Currently we don't write more than one message
at a time, primarily due to limitations in
the Poco sockets in the unit-tests, which
have a hard time consuming multiple WS frames
with a single poll (subsequent calls to poll
doesn't enter signalled state until new data
arrives, possibly because the data is read and
buffered internally, making the whole scheme
of using poll unreliable and meaningless).
Change-Id: Ic2e2cf1babfb5ab4116efd93f392977ba234d92b
Signed-off-by: Ashod Nakashian <ashod.nakashian@collabora.co.uk>
* and fix error reporting: we do not abort every time when loolforkit
is invoked with incorrect user name
* and better readability of the conditions
Signed-off-by: Andras Timar <andras.timar@collabora.com>
Change-Id: Idc9db40c00d41c95160db130eb324c487f14de17
Cleans up some of the conversions and implicity
unit in integral types.
Change-Id: I79f35b92f8f631894e55bdb39851b050870fce96
Signed-off-by: Ashod Nakashian <ashod.nakashian@collabora.co.uk>
Doing this check in loolmount complicates using it
from the multiple places we do. It hardly adds
any advantage, since we already perform the lool-user
UID check in forkit, at startup.
Change-Id: I6fa1b546663a6a3a3816d4d637b4acae1d09fccb
See also f1be65668c
systemplate files should not be writable by lool user
Change-Id: I5684248d3d4b4b0ba56f8c5ab490a6e7df0e0038
Reviewed-on: https://gerrit.libreoffice.org/c/online/+/98069
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Tested-by: Jenkins
Reviewed-by: Andras Timar <andras.timar@collabora.com>
loolmount now works and supports mounting and
unmounting, plus numerous improvements,
refactoring, logging, etc.. When enabled,
binding improves the jail setup time by anywhere
from 2x to orders of magnitude (in docker, f.e.).
A new config entry mount_jail_tree controls
whether mounting is used or the old method of
linking/copying of jail contents. It is set to
true by default and falls back to linking/copying.
A test mount is done when the setting is enabled,
and if mounting fails, it's disabled to avoid noise.
Temporarily disabled for unit-tests until we can
cleanup lingering mounts after Jenkins aborts our
build job. In a future patch we will have mount/jail
cleanup as part of make.
The network/system files in /etc that need frequent
refreshing are now updated in systemplate to make
their most recent version available in the jails.
These files can change during the course of loolwsd
lifetime, and are unlikely to be updated in
systemplate after installation at all. We link to
them in the systemplate/etc directory, and if that
fails, we copy them before forking each kit
instance to have the latest.
This reworks the approach used to bind-mount the
jails and the templates such that the total is
now down to only three mounts: systemplate, lo, tmp.
As now systemplate and lotemplate are shared, they
must be mounted as readonly, this means that user/
must now be moved into tmp/user/ which is writable.
The mount-points must be recursive, because we mount
lo/ within the mount-point of systemplate (which is
the root of the jail). But because we (re)bind
recursively, and because both systemplate and
lotemplate are mounted for each jails, we need to
make them unbindable, so they wouldn't multiply the
mount-points for each jails (an explosive growth!)
Contrarywise, we don't want the mount-points to
be shared, because we don't expect to add/remove
mounts after a jail is created.
The random temp directory is now created and set
correctly, plus many logging and other improvements.
Change-Id: Iae3fda5e876cf47d2cae6669a87b5b826a8748df
Reviewed-on: https://gerrit.libreoffice.org/c/online/+/92829
Tested-by: Jenkins
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Reviewed-by: Ashod Nakashian <ashnakash@gmail.com>
The tokenizer(s) are more generic than the protocol
logic, and are used from contexts that don't involve
the protocol as such.
Change-Id: Ie8c256bf11a91e466bff794021f41603c9596a7f
More readable and typically more efficient.
Change-Id: I9bd5bfc91f4ac255bb8ae0987708fb8b56b398f8
Reviewed-on: https://gerrit.libreoffice.org/c/online/+/95285
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Jenkins
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
Weak pointers can be null and must be
checked before using. This fixes at least
one segfault and prevents a number of others.
Also, minimizes locking of weak pointers
in the message handlers.
Change-Id: I306501c26c3441d7bd6812d51fa17e7356126f32
Reviewed-on: https://gerrit.libreoffice.org/c/online/+/92828
Reviewed-by: Michael Meeks <michael.meeks@collabora.com>
Tested-by: Jenkins CollaboraOffice <jenkinscollaboraoffice@gmail.com>
