These were previously reporting leaks due to various allocations not
getting cleaned up during the stack unwind, but at the current state of
the transition toward IntrusivePtr usage, theses tests no longer leak.
* origin/topic/jsiwek/gh-857-rebuild-tables-on-record-redef:
Initialize RecordVal default fields when redef'd
GH-857: fix redefining record types used to index tables
Change RecordVals to get resized at time of RecordType redef
* 'intrusive_ptr' of https://github.com/MaxKellermann/zeek: (32 commits)
Scope: store IntrusivePtr in `local`
Scope: pass IntrusivePtr to AddInit()
DNS_Mgr: use class IntrusivePtr
Scope: use class IntrusivePtr
Attr: use class IntrusivePtr
Expr: check_and_promote_expr() returns IntrusivePtr
Frame: use class IntrusivePtr
Val: RecordVal::LookupWithDefault() returns IntrusivePtr
Type: RecordType::FieldDefault() returns IntrusivePtr
Val: TableVal::Delete() returns IntrusivePtr
Type: base_type() returns IntrusivePtr
Type: init_type() returns IntrusivePtr
Type: merge_types() returns IntrusivePtr
Type: use class IntrusivePtr in VectorType
Type: use class IntrusivePtr in EnumType
Type: use class IntrusivePtr in FileType
Type: use class IntrusivePtr in TypeDecl
Type: make TypeDecl `final` and the dtor non-`virtual`
Type: use class IntrusivePtr in TypeType
Type: use class IntrusivePtr in FuncType
...
* The compression capability was incorrectly set to 0x0004 instead of 0x0003
* The padding was 4-byte instead of 8-byte aligned and also the spec.
does not strictly require the padding for the last item in the list.
* Add a default case to handle parsing of unknown context types.
This change tracks all TableVals created at parse-time whose index
depends on a given RecordType. Should that RecordType be redef'd, those
TableVals are immediately rebuilt such that they are valid to
subsequently use in either parse-time initializations or eventually in
any arbitrary run-time expression.
This lets packet source plugins implement handling of hardware checksum offloading, if available. Setting the flags will skip the internal checksumming for either layer 2 and/or layer 3.
* topic/johanna/gh-744-cache-the-heck-out-of-these-certs:
X509 caching API change: callback gets entry directly
Address feedback
Small default updates & external baselines.
X509 caching: small API changes, tests & test updates
Certificate caching - now working in principle.
First (sadly not completely working) approach at caching X509 certificates
Changed some configuration defaults to potentially more same values.
The callback function is now a hook to allow costomization of the events
that are raised.
Tests now exist. Test baselines are updated.
Added a comment to clarify comparison of EOF packet length to 13.
* origin/topic/vladg/gh-843:
Update baselines
Made additional MySQL fixes.
Add support to MySQL for deprecation of EOF packets.
Whitespace cleanup & fixes
Fix EOF detection in the MySQL protocol analyzer.
I moved the replay function to a callback - which now means that the replayed
functions are called before file_state remove. Which makes this virtually
identical with the events raised by the core.
Currently this is mostly missing tests, leak-testing and performance-tests.
1) There are a couple more places where the new protocol uses and OK
packet instead of the deprecated EOF.
2) With > 255 results, we could end up in an situation where the uint8
sequence number would wrap, and we'd naively think it'd be a new
handshake.
Now, we track the previous sequence number, and assume overflow if it
was 255 previously and 0 now.
We also reset the previous sequence number to 0 in various packets
that we'd expect at the end of other commands.
This approach mostly relies on script-level changes. In scriptland, we track
which certificates should be cached - and also perform the caching and the
replaying of events there.
This actually is probably nearly functional - the problem is the fact that now
the event ordering is wrong - and file_state_remove is called before the x509
events are thrown.
The fix probably is to move to a callback from the core - which can execute
things in the right order again. (Or just write the whole event-raising inside
the core - but that is both less flexible and a lot more cumbersome).
