aboutsummaryrefslogtreecommitdiffstats
path: root/lib
AgeCommit message (Collapse)Author
2015-03-30Merge branch 'egil/maps/tests'Björn-Egil Dahlberg
* egil/maps/tests: debugger: Strengthen Maps tests compiler: Strengthen Maps tests erts: Strengthen Maps tests
2015-03-30Merge branch 'maint'Dan Gudmundsson
* maint: observer: Fix crash in process_window
2015-03-30Merge branch 'dgud/observer/fix-atom-tostr/OTP-12634' into maintDan Gudmundsson
* dgud/observer/fix-atom-tostr/OTP-12634: observer: Fix crash in process_window
2015-03-27observer: Fix crash in process_windowDan Gudmundsson
Links can be to named processes, convert atom to list
2015-03-27Merge branch 'maint'Anders Svensson
2015-03-27Merge branch 'maint'Sverker Eriksson
Conflicts: lib/stdlib/src/stdlib.appup.src
2015-03-27Merge branch 'anders/diameter/hardening/OTP-12628' into maintAnders Svensson
* anders/diameter/hardening/OTP-12628: Remove potentially large error reason in call to diameter_lib:log/4 Limit FQDN in DiameterURI to 255 octets Limit DiameterURI ports to 0-65535 digits on decode Add service_opt() incoming_maxlen Add guard to reject {spawn_opt, false} as transport/service_opt()
2015-03-27Remove potentially large error reason in call to diameter_lib:log/4Anders Svensson
The function is intended to be traced on, to see abnormalities (mostly) without producing excessive output. In the case of decode failure, the error reason can be things like {badmatch, HugeBinary}. Missed in commit 0058430.
2015-03-27Merge branch 'dgud/mnesia/try-catch'Dan Gudmundsson
* dgud/mnesia/try-catch: mnesia: Replace catch with try-catch
2015-03-27mnesia: Replace catch with try-catchDan Gudmundsson
Avoids building stacktraces where it is not needed and do not mask errors, i.e. only catch the relevant classes in each try.
2015-03-27Merge branch 'dgud/wx/prefer-wxMSW-3/OTP-12632'Dan Gudmundsson
* dgud/wx/prefer-wxMSW-3/OTP-12632: wx: Fix variable cursor constants wx: On Mac keyboards use meta field as raw_ctrl wx: Prefer wxWidgets-3.0. branch on windows
2015-03-27wx: Fix variable cursor constantsDan Gudmundsson
Fix cursor constants that changes with OS and wxWidgets releases.
2015-03-27wx: On Mac keyboards use meta field as raw_ctrlDan Gudmundsson
On mac and wxWidgets-3.0 the crtlDown field in keyboard and mouse events are set when command button is pressed. The ctrl key on mac keyboard is in a field called raw_control as this would be backward incompatible to add we reuse the metaDown field as indicator when the ctrl keyboard button is pressed.
2015-03-27wx: Prefer wxWidgets-3.0. branch on windowsDan Gudmundsson
Since no bug fixes are done anymore on the 2.8 branch it is time to switch the preferred version to 3.0 branch.
2015-03-27Merge branch 'dgud/observer/allocator-graphs/OTP-12631'Dan Gudmundsson
* dgud/observer/allocator-graphs/OTP-12631: observer: Fix menu code (for mac) observer: Add a new tab memory allocator charts
2015-03-27observer: Fix menu code (for mac)Dan Gudmundsson
Have been broken since: 6fe02e14d observer: Add SASL log view for processes
2015-03-27observer: Add a new tab memory allocator chartsDan Gudmundsson
Easier to the see memory usage and utilization as charts instead of values that is updating during runtime. Remove from system page and create a separate allocator tab.
2015-03-27debugger: Strengthen Maps testsBjörn-Egil Dahlberg
2015-03-27compiler: Strengthen Maps testsBjörn-Egil Dahlberg
2015-03-27Limit FQDN in DiameterURI to 255 octetsAnders Svensson
As for the port number in the parent commit, a FQDN can't be arbitrarily long, at most 255 octets. Make decode fail if it's more.
2015-03-27Limit DiameterURI ports to 0-65535 digits on decodeAnders Svensson
A port number is a 16-bit integer, but the regexp used to parse it in commit 1590920 slavishly followed the RFC 6733 grammar in matching an arbitrary number of digits. Make decode fail if it's anything more than 5, to avoid doing erlang:list_to_integer/1 on arbitrarily large lists. Also make it fail if the resulting integer is outside of the expected range.
2015-03-27Add service_opt() incoming_maxlenAnders Svensson
To bound the length of incoming messages that will be decoded. A message longer than the specified number of bytes is discarded. An incoming_maxlen_exceeded counter is incremented to make note of the occurrence. The motivation is to prevent a sufficiently malicious peer from generating significant load by sending long messages with many AVPs for diameter to decode. The 24-bit message length header accomodates (16#FFFFFF - 20) div 12 = 1398099 Unsigned32 AVPs for example, which the current record-valued decode is too slow with in practice. A bound of 16#FFFF bytes allows for 5461 small AVPs, which is probably more than enough for the majority of applications, but the default is the full 16#FFFFFF.
2015-03-26stdlib: Support upgrading from and to stdlib-2.3Björn Gustavsson
2015-03-26Merge branch 'egil/maps/refactor-tagscheme/OTP-12585'Björn-Egil Dahlberg
* egil/maps/refactor-tagscheme/OTP-12585: erts: Refactor Map - use multiple values ESTACK_PUSHN erts: GC needs the size even if the frag is not referenced Revert "hipe: Handle separate hashmap tag correctly" erts: Combine flat and hash maps under one unifying tag
2015-03-26Add guard to reject {spawn_opt, false} as transport/service_opt()Anders Svensson
It was possible to configure the option, but doing so caused the service to fail when starting a watchdog process: {function_clause, [{diameter_service,'-spawn_opts/1-lc$^0/1-0-', [false], [{file,"base/diameter_service.erl"},{line,846}]}, {diameter_service,start,5, [{file,"base/diameter_service.erl"},{line,820}]}, {diameter_service,start,3, [{file,"base/diameter_service.erl"},{line,782}]}, {diameter_service,handle_call,3, [{file,"base/diameter_service.erl"},{line,385}]}, {gen_server,try_handle_call,4,[{file,"gen_server.erl"},{line,607}]}, {gen_server,handle_msg,5,[{file,"gen_server.erl"},{line,639}]}, {proc_lib,init_p_do_apply,3,[{file,"proc_lib.erl"},{line,237}]}]} Tests for the option in the config suite were also missing. Bungled in commit 78b3dc6.
2015-03-26Merge branch 'maint'Peter Andersson
2015-03-26Merge branch 'peppe/common_test/failing_telnet_tests' into maintPeter Andersson
* peppe/common_test/failing_telnet_tests: Add documentation and make get_data behaviour backwards compatible Introduce polling feature in ct_telnet Fix problem with invalid timeouts because of truncated floats OTP-12627
2015-03-26Add documentation and make get_data behaviour backwards compatiblePeter Andersson
2015-03-25Merge branch 'bjorn/doc'Björn Gustavsson
* bjorn/doc: cerl_trees: Fix incorrect EDoc reference to the cerl module cerl: Correct incorrect EDoc references
2015-03-25Merge branch 'erland/OTP18/inets/time_api/OTP-12441'Erland Schönbeck
* erland/OTP18/inets/time_api/OTP-12441: inets: Add new module inets_time_compat with new time API The new module is backwards compatible.
2015-03-25Merge branch 'erland/ssh/OTP18/time_api/OTP-12444'Erland Schönbeck
* erland/ssh/OTP18/time_api/OTP-12444: ssh: Use new time API
2015-03-25Revert "hipe: Handle separate hashmap tag correctly"Björn-Egil Dahlberg
This reverts commit 4dc5f194b0a5a6f052763bb7494b0458f98d96ca.
2015-03-24Introduce polling feature in ct_telnetPeter Andersson
2015-03-24inets: Add new module inets_time_compat with new time APIErland Schönbeck
The new module is backwards compatible.
2015-03-24cerl_trees: Fix incorrect EDoc reference to the cerl moduleBjörn Gustavsson
2015-03-24Merge branch 'maint'Anders Svensson
2015-03-24vsn -> 1.9Anders Svensson
2015-03-24Update appup for 17.5Anders Svensson
Required load order by ticket. - OTP-11492, answer messages discarded - OTP-12415, retransmission failure - OTP-12475, grouped AVP decode - OTP-12543, no requests after DPR none - OTP-12412, shutdown issues diameter_lib diameter_service - OTP-12428, transport_opt() pool_size diameter_lib diameter_service diameter, diameter_config diameter_{tcp,sctp} diameter, diameter_config - OTP-12439, new time api in Erlang/OTP 18 diameter_lib diameter_{config,peer,reg,service,session,stats,sync,watchdog,sctp} - OTP-11952, service_opt() decode_string - OTP-12589, DiameterURI encode/decode diameter_{capx,codec,peer} diameter_types diameter_traffic diameter_{service,peer_fsm} diameter_watchdog diameter, diameter_config - OTP-12542, DPR with diameter:call/4 diameter_{peer_fsm,watchdog} diameter, diameter_config - OTP-12609, transport_opt() dpr_timeout diameter_peer_fsm diameter, diameter_config
2015-03-24Minor doc fixAnders Svensson
2015-03-24Merge branch 'maint'Anders Svensson
2015-03-24Merge branch 'anders/diameter/dpr/OTP-12609' into maintAnders Svensson
* anders/diameter/dpr/OTP-12609: Discard incoming/outgoing requests after incoming DPR Add transport_opt() dpr_timeout Be lenient with errors in incoming DPR
2015-03-24Adapt to changed DiameterURI defaults in RFC 6733Anders Svensson
Despite claims of full backwards compatibility, the text of RFC 6733 changes the interpretation of unspecified values in a DiameterURI. In particular, 3588 says that the default port and transport are 3868 and sctp respectively, while 6733 says it's either 3868/tcp (aaa) or 5658/tcp (aaas). The 3588 defaults were used regardless, but now use them only if the common dictionary is diameter_gen_base_rfc3588. The 6733 defaults are used otherwise. This kind of change in the standard can lead to interop problems, since a node has to know which RFC its peer is following to know that it will properly interpret missing URI components. Encode of a URI includes all components to avoid such confusion. That said, note that the defaults in the diameter_uri record have *not* been changed. This avoids breaking code that depends on them, but the risk is that such code sends inappropriate values. The record defaults may be changed in a future release, to force values to be explicitly specified.
2015-03-24Reject transport=udp;protocol=diameter at DiameterURI encodeAnders Svensson
Both RFC 3588 and 6733 disallow the combination. Make its encode fail.
2015-03-24Merge branch 'anders/diameter/string_decode/OTP-11952' into maintAnders Svensson
* anders/diameter/string_decode/OTP-11952: Let examples override default service options Set {restrict_connections, false} in example server Set {string_decode, false} in examples Test {string_decode, false} in traffic suite Add service_opt() string_decode Strip potentially large terms when sending outgoing Diameter messages Improve language consistency in diameter(1)
2015-03-24Merge branch 'anders/diameter/route_record/OTP-12551' into maintAnders Svensson
* anders/diameter/route_record/OTP-12551: Fix ordering of AVPs in relayed messages
2015-03-24Let examples override default service optionsAnders Svensson
To make them a bit more flexible. Can now do things like this: server:start([{'Product-Name', "Bob"}]), server:listen({tcp, [{capx_timeout, 2000}]}) Beware that the latter is completely different from this: server:listen(tcp, [{capx_timeout, 2000}])
2015-03-24Set {restrict_connections, false} in example serverAnders Svensson
Since there's no reason to reject a client that wants to establish multiple connections, given that diameter can handle it.
2015-03-24Set {string_decode, false} in examplesAnders Svensson
So as to do what's now recommended in diameter(1), in the grandparent commit.
2015-03-24Test {string_decode, false} in traffic suiteAnders Svensson
By adding string decode or not in the server or client as another combination. Run all traffic cases in parallel: remove the sequential tests. Common test seems unable to deal with {group, X, [parallel]} within a group.
2015-03-24Add service_opt() string_decodeAnders Svensson
To control whether stringish Diameter types are decoded to string or left as binary. The motivation is the same as in the parent commit: to avoid large strings being copied when incoming Diameter messages are passed between processes; or *if* in the case of messages destined for handle_request and handle_answer callbacks, since these are decoded in the dedicated processes that the callbacks take place in. It would be possible to do something about other messages without requiring an option, but disabling the decode is the most effective. The value is a boolean(), true being the default for backwards compatibility. Setting false causes both diameter_caps records and decoded messages to contain binary() in relevant places that previously had string(): diameter_app(3) callbacks need to be prepared for the change. The Diameter types affected are OctetString and the derived types that can contain arbitrarily large values: OctetString, UTF8String, DiameterIdentity, DiameterURI, IPFilterRule, and QoSFilterRule. Time and Address are unaffected. The DiameterURI decode has been redone using re(3), which both simplifies and does away with a vulnerability resulting from the conversion of arbitrary strings to atom. The solution continues the use and abuse of the process dictionary for encode/decode purposes, last seen in commit 0f9cdba.