Only 1% build time speedup, but still, it declutters the headers a bit.
Before this patch:
2565.17user 141.83system 2:25.46elapsed 1860%CPU (0avgtext+0avgdata 1489076maxresident)k
72576inputs+9130920outputs (1667major+49400430minor)pagefaults 0swaps
After this patch:
2537.19user 142.94system 2:26.90elapsed 1824%CPU (0avgtext+0avgdata 1434268maxresident)k
16240inputs+8887152outputs (1931major+48728888minor)pagefaults 0swaps
The Zeek code base has very inconsistent #includes. Many sources
included a few headers, and those headers included other headers, and
in the end, nearly everything is included everywhere, so missing
#includes were never noticed. Another side effect was a lot of header
bloat which slows down the build.
First step to fix it: in each source file, its own header should be
included first to verify that each header's includes are correct, and
none is missing.
After adding the missing #includes, I replaced lots of #includes
inside headers with class forward declarations. In most headers,
object pointers are never referenced, so declaring the function
prototypes with forward-declared classes is just fine.
This patch speeds up the build by 19%, because each compilation unit
gets smaller. Here are the "time" numbers for a fresh build (with a
warm page cache but without ccache):
Before this patch:
3144.94user 161.63system 3:02.87elapsed 1808%CPU (0avgtext+0avgdata 2168608maxresident)k
760inputs+12008400outputs (1511major+57747204minor)pagefaults 0swaps
After this patch:
2565.17user 141.83system 2:25.46elapsed 1860%CPU (0avgtext+0avgdata 1489076maxresident)k
72576inputs+9130920outputs (1667major+49400430minor)pagefaults 0swaps
No need to create a record introspection table each time when all
the required information can be obtained directly in the RecordVal
and RecordType objects. Besides the additional overhead, using such
a table will re-order the fields arbitrarily instead of using the
order in which they're defined.
* origin/topic/timw/171-deprecation-warnings:
GH-171: support warning messages alongside deprecated attributes
Made a minor tweak to give a better error message when using
&deprecated= with something other than a string literal expression.
Fixed a few small bugs - Modifiable had an uninitialized member and the
Registry looped over a map while deleting elements from it.
Fixes GH-319
* remotes/origin/topic/robin/gh59-when:
Renaming src/StateAccess.{h,cc} to src/Notifier.{h,cc}.
Clean up new code.
Remove MutableVal class.
Redo API for notifiers.
Remove most of MutableVal (but not the class itelf yet)
Remove enum Opcode.
Remove StateAccess class.
Redo NotfifierRegistry to no longer rely on StateAccess.
Add new test for when-statement watching global variables.
Couple of compile fixes.
It turns out that SetType was missing a ShallowClone implementation.
Which meant that when a SetType was cloned, a TableType was received
(which has one less member). Which resulted in invalid memory accesses
when using the clone.
Thank you valgrind :)
This is needed to track name changes for the documentation.
With this things, which do not need val-cloning, generally seem to work
again. There are a whole bunch of test failures at the moment.
Note - this compiles, but you cannot run Bro anymore - it crashes
immediately with a 0-pointer access. The reason behind it is that the
required clone functionality does not work anymore.
Majority of PLists are now created as automatic/stack objects,
rather than on heap and initialized either with the known-capacity
reserved upfront or directly from an initializer_list (so there's no
wasted slack in the memory that gets allocated for lists containing
a fixed/known number of elements).
Added versions of the ConnectionEvent/QueueEvent methods that take
a val_list by value.
Added a move ctor/assign-operator to Plists to allow passing them
around without having to copy the underlying array of pointers.
This commit marks (hopefully) ever one-parameter constructor as explicit.
It also uses override in (hopefully) all circumstances where a virtual
method is overridden.
There are a very few other minor changes - most of them were necessary
to get everything to compile (like one additional constructor). In one
case I changed an implicit operation to an explicit string conversion -
I think the automatically chosen conversion was much more convoluted.
This took longer than I want to admit but not as long as I feared :)
Instead of having a big event, that tries to parse all the data into a
huge datastructure, we do the more common thing and use a series of
smaller events to parse requests and responses.
The new events are:
ocsp_request -> raised for an ocsp request, giving version and requestor
ocsp_request_certificate -> raised n times per request, once per cert
ocsp_response_status -> raised for each ocsp response, giving status
ocsp_response_bytes -> raised for each ocsp response with information
ocsp_response_certificate -> raised for each cert in an ocsp response
While scripts are parsed, a warning is raised for each usage of an
identifier marked as &deprecated. This also works for BIFs.
Addresses BIT-924, BIT-757.
chain. This means that it is now possible to get information about the
root-certificates that were used to secure a connection.
Intermediate commit before changing the script interface again.
addresses BIT-953, BIT-760
This changes the internal type that is used to signal that a vector
is unspecified from any to void.
I tried to verify that the behavior of Bro is still the same. After
a lot of playing around, I think everything still should worl as before.
However, it might be good for someone to take a look at this.
addresses BIT-1144
- Internals: move type alias table to private static BroType member.
- Sphinx extension: now uses absolute path to bro binary.
- reST ouput formatting: remove "param" from function desriptions
and change package overview docs so script link+summaries render
consistently.
BIT-1098
* origin/topic/jsiwek/broxygen:
Fix Broxygen-related compile errors.
Add a Broxygen coverage test.
Internal Broxygen organization/documentation/polish.
Add unit tests for Broxygen config file targets.
Change Broxygen config file format.
Broxygen doc-related test updates. Fix two regressions.
A couple documentation fixes.
Integrate new Broxygen functionality into Sphinx.
Implement majority of Broxygen features delegated to Bro.
Broxygen can now read a config file specifying particular targets.
Remove unneeded Broxygen comments in scan.bro.
Replace safe_basename/safe_dirname w/ SafeBasename/SafeDirname.
Add BIF interface for retrieving comments/docs.
Quick optimization to Broxygen doc gathering.
Flesh out Broxygen doc-gathering skeleton.
Refactor search_for_file() util function.
Initial skeleton of new Broxygen infrastructure.
Broccoli clients can only recv events w/ vectors for now.
Also changed ordering of Bro type tag enum -- the addition of opaque
types changed the value of the vector type, making broccoli.h's
definition out of sync. Probably could have just changed broccoli's
definition, but seems more correct to go back to using the same value
for vectors as they were before opaques. It's also better in case
there's some other location I'm not aware of where the values are
replicated.
