aboutsummaryrefslogtreecommitdiffstats
path: root/erts/emulator/beam/copy.c
AgeCommit message (Collapse)Author
2019-02-22erts: Add ERL_NODE_BOOKKEEP to node tables refcLukas Larsson
2018-10-24Extend the sharing-preserving routines to optionally copy literalsBjörn Gustavsson
In the implementation of the zero-copying term storage, we want to preserve sharing, but not copy literals because the modules holding the literals could be unloaded under our feet.
2018-06-18Update copyright yearHenrik Nord
2018-01-29More ERTS_RESTRICT usageSverker Eriksson
2018-01-19erts: Optimize move_cons & move_boxedSverker Eriksson
Replace double pointer with return that can mostly be ignored. Use restrict pointers.
2017-07-17erts: Replace usage of all erts_smp prefixes to just ertsLukas Larsson
2017-05-04Update copyright yearRaimo Niskanen
2017-04-12erts: Introduce struct binary_internalsSverker Eriksson
to replace macro ERTS_INTERNAL_BINARY_FIELDS as header in Binary and friends.
2017-03-28erts: Fix faulty assert for refs in copy sharingBjörn-Egil Dahlberg
2017-02-28erts: Refactor MOVE_CONS to inline functionBjörn-Egil Dahlberg
2017-02-28erts: Refactor MOVE_BOXED to inline functionBjörn-Egil Dahlberg
2017-02-06Implement magic referencesRickard Green
Magic references are *intentionally* indistinguishable from ordinary references for the Erlang software. Magic references do not change the language, and are intended as a pure runtime internal optimization. An ordinary reference is typically used as a key in some table. A magic reference has a direct pointer to a reference counted magic binary. This makes it possible to implement various things without having to do lookups in a table, but instead access the data directly. Besides very fast lookups this can also improve scalability by removing a potentially contended table. A couple of examples of planned future usage of magic references are ETS table identifiers, and BIF timer identifiers. Besides future optimizations using magic references it should also be possible to replace the exposed magic binary cludge with magic references. That is, magic binaries that are exposed as empty binaries to the Erlang software.
2017-02-06Merge branch 'maint'Rickard Green
* maint: Atomic reference count of binaries also in non-SMP Conflicts: erts/emulator/beam/erl_fun.c
2017-02-06Atomic reference count of binaries also in non-SMPRickard Green
NIF resources was not handled in a thread-safe manner in the runtime system without SMP support. As a consequence of this fix, the following driver functions are now thread-safe also in the runtime system without SMP support: - driver_free_binary() - driver_realloc_binary() - driver_binary_get_refc() - driver_binary_inc_refc() - driver_binary_dec_refc()
2016-08-01erts: Refactor literal purge area argumentsBjörn-Egil Dahlberg
2016-08-01erts: Remove the need for copying of literalsBjörn-Egil Dahlberg
* Literals are not copied between processes for messages or spawn Increases performance of message sent and processes spawned when literals are involved in messages or arguments.
2016-03-15update copyright-yearHenrik Nord
2016-02-24Merge branch 'master' into sverk/master/halt-INT_MINSverker Eriksson
2016-02-24erts: Change erl_exit into erts_exitSverker Eriksson
This is mostly a pure refactoring. Except for the buggy cases when calling erlang:halt() with a positive integer in the range -(INT_MIN+2) to -INT_MIN that got confused with ERTS_ABORT_EXIT, ERTS_DUMP_EXIT and ERTS_INTR_EXIT. Outcome OLD erl_exit(n, ) NEW erts_exit(n, ) ------- ------------------- ------------------------------------------- exit(Status) n = -Status <= 0 n = Status >= 0 crashdump+abort n > 0, ignore n n = ERTS_ERROR_EXIT < 0 The outcome of the old ERTS_ABORT_EXIT, ERTS_INTR_EXIT and ERTS_DUMP_EXIT are the same as before (even though their values have changed).
2015-11-18Refactor sharing preserved copy flagsBjörn-Egil Dahlberg
The TMPBUF option is no longer needed due to is_literal test and NONE was only used for initial debugging. So we remove the entire option.
2015-11-17Do not use GCC extensions in copyBjörn-Egil Dahlberg
2015-11-17Use copy literal range check in message passing and purgingBjörn-Egil Dahlberg
2015-11-17Copy literals in copy sharingBjörn-Egil Dahlberg
2015-11-17Refactor copy sharingBjörn-Egil Dahlberg
2015-11-17Add support for HAMT maps in preserved copyBjörn-Egil Dahlberg
2015-11-17Fix Map preserved sharing copy implementationBjörn-Egil Dahlberg
The Map implementation has changed since initial preserved copy implementation.
2015-11-17Fix Halfword removalBjörn-Egil Dahlberg
Halfword is no longer present in the runtime system.
2015-11-17Fix internal stacksBjörn-Egil Dahlberg
The internal stacks has changed between releases.
2015-11-17Add support for maps in preserved copyNikolaos S. Papaspyrou
2015-11-17Add -debug +vc flag for debuging SHCOPYNikolaos S. Papaspyrou
This is very verbose, you have been warned. It should work with the copy-spy.py script, which may be a bit outdated.
2015-11-17Add machinery to enable SHCOPY dynamicallyNikolaos S. Papaspyrou
This commit is just for debugging purposes, will probably be reverted. It comes with a the erts_debug:copy_shared/1 BIF. If SHCOPY_DISABLE is defined, SHCOPY starts disabled and is dynamically enabled the first time that the BIF is called.
2015-11-17Add all the main machineryNikolaos S. Papaspyrou
Add functions size_shared, copy_shared_calculate and copy_shared_perform. Add the infrastructure for making these communicate with each other. Add debug information to other places in the VM, to watch interaction with the sharing-preserving copy. CAUTION: If you define the SHCOPY_DEBUG macro (after SHCOPY is actually used in the VM) and make the whole OTP, there will be a lot of debugging messages during make (it will also be enabled in erlc). You have been warned...
2015-11-12Fragmented young heap generation and off_heap_message_queue optionRickard Green
* The youngest generation of the heap can now consist of multiple blocks. Heap fragments and message fragments are added to the youngest generation when needed without triggering a GC. After a GC the youngest generation is contained in one single block. * The off_heap_message_queue process flag has been added. When enabled all message data in the queue is kept off heap. When a message is selected from the queue, the message fragment (or heap fragment) containing the actual message is attached to the youngest generation. Messages stored off heap is not part of GC.
2015-11-12Introduce literal tagRickard Green
2015-06-24erts: Remove halfword BINARY RELsBjörn-Egil Dahlberg
* ERTS_GET_BINARY_BYTES_REL * ERTS_GET_REAL_BIN_REL
2015-06-24erts: Reinstate copy_object over-allocation optimizationBjörn-Egil Dahlberg
2015-06-24erts: Remove halfword basic relative heap operationsBjörn-Egil Dahlberg
2015-06-24erts: Remove HALFWORD_HEAP definitionBjörn-Egil Dahlberg
2015-06-18Change license text to APLv2Bruce Yinhe
2015-06-15erts: Remove hashmap probabilistic heap overestimationSverker Eriksson
by adding a dynamic heap factory. "binary_to_term" is now a hybrid solution with both a call to decoded_size() to calculate needed heap space AND possible dynamic allocation of more heap space if needed for big maps. The heap size returned from decoded_size() is guaranteed to be sufficient for all term heap data except for hashmap nodes. All hashmap nodes are created at the end of dec_term() by invoking the heap factory interface that may allocate more heap space on process heap or in fragments. With this commit it is no longer guaranteed that a message is confined to only one heap fragment.
2015-05-07erts: Fix copy shallow for large MapsBjörn-Egil Dahlberg
There is no need to take special care of Maps at all since header_arity(hdr) will take care of the normal case via its Map handling.
2015-03-25erts: Combine flat and hash maps under one unifying tagBjörn-Egil Dahlberg
2015-03-13erts: Fix typo in copy_struct for halfword emulatorBjörn-Egil Dahlberg
2015-03-12erts: Refactor maps naming conventionSverker Eriksson
flatmap: Small map hashmap: Large map map: flatmap or hashmap
2015-03-12erts: Remove erl_hashmap.[ch] filesBjörn-Egil Dahlberg
2015-03-12Initial Persistent HAMT - Map frameworkBjörn-Egil Dahlberg
Conflicts: erts/emulator/Makefile.in erts/emulator/beam/bif.tab erts/emulator/beam/erl_gc.c erts/emulator/beam/erl_gc.h erts/emulator/beam/erl_printf_term.c erts/emulator/beam/erl_term.c erts/emulator/beam/erl_term.h
2014-10-24erts: Make hipe_{un}reserve_beam_trap_frame INLINESverker Eriksson
2014-03-25Merge branch 'paulgray/dtrace_snprintf_fix/OTP-11816'Lukas Larsson
* paulgray/dtrace_snprintf_fix/OTP-11816: Pass full strings to DTrace probes
2014-03-12Pass full strings to DTrace probesMichal Ptaszek
Whenever string is passed as an argument to a DTrace probe, its length should be properly computed. Until now in order to get length of the input buffer size_of(char *) was used - which evalutes to 4 or 8 (depending on the architecture). To get a proper length, size_of(DTRACE_CHARBUF_NAME(buffer_name)) should be used.
2014-01-28erts: Initial Map instructions, type and structureBjörn-Egil Dahlberg