| Age | Commit message (Collapse) | Author |
|---|
|
Consider this function and its corresponding BEAM code:
foo(Map, Key) ->
Val = case Map of
#{Key:=Val0} -> Val0;
_ -> default
end,
bar(1, 2, Val).
{label,2}.
{test,is_map,{f,3},[{x,0}]}.
{get_map_elements,{f,3},{x,0},{list,[{x,1},{x,0}]}}.
^^^^^
{jump,{f,4}}.
{label,3}.
{move,{atom,default},{x,0}}.
^^^^^
{label,4}.
{move,{integer,2},{x,1}}.
{move,{x,0},{x,2}}.
^^^^^
{move,{integer,1},{x,0}}.
{call_only,3,{f,6}}.
Note that the value of the variable `Val` will first be
placed in `{x,0}` and then moved to `{x,2}` where it needs
to be when calling the `bar/3` function.
The reason for the extra `move` instruction is that the
register allocator picks the lowest numbered available register
when choosing a register to put a variable in. In this case,
`{x,0}` will be chosen.
If we only could give a hint to the register allocator that
it would be better to put `Val` in `{x,2}`, the extra `move`
would disappear:
{label,2}.
{test,is_map,{f,3},[{x,0}]}.
{get_map_elements,{f,3},{x,0},{list,[{x,1},{x,2}]}}.
{jump,{f,4}}.
{label,3}.
{move,{atom,default},{x,2}}.
{label,4}.
{move,{integer,2},{x,1}}.
{move,{integer,1},{x,0}}.
{call_only,3,{f,6}}.
There already is an existing sub pass (`reserve_regs`) in
`beam_ssa_pre_codegen` that among things tries to give the register
allocator hints that some variables should be placed in specific
registers, if possible. However, the existing hinting mechanism is
limited, essentially only working within a single SSA block.
This commit extends the hinting mechanism, allowing hints to be passed
across SSA blocks, eliminating `move` instructions and register
shuffling in many places. (494 modules out of a sample of 1236 modules
were changed by this commit.)
|
|
Use lists:splitwith/2 instead of lists:partition/2 for splitting out
phi nodes. Since phi nodes are always the first instructions in a
block, the result will be the same, but splitwith/2 is faster.
|
|
|
|
Conflicts:
lib/ssl/src/tls_connection.erl
|
|
* lukas/erts/fix_inet_multitimer_cleanup/OTP-15536:
erts: Fix cleanup of the inet MultiTimer
|
|
* ingela/ssl/continue-optimize/OTP-15445:
ssl: If possible assemble several received application data records
|
|
|
|
* sverker/fix-since-clauses/OTP-15460:
Fix "since" for all multi clause functions
|
|
|
|
Conflicts:
lib/ssl/src/ssl_connection.erl
lib/ssl/src/ssl_connection.hrl
lib/ssl/src/tls_connection.erl
|
|
Specify that mnesia:start/0 is async
|
|
We want to decrease the size of the outer state tuple, and gain
ease of understanding by better grouping. This is the first step
of creating a hs_env (handshake environment) part of the state.
This change will be performed gradually to reduce merge conflicts
complexity and risk of introducing errors.
|
|
|
|
|
|
|
|
* ingela/ssl/test-cuddle:
ssl: Correct test input
|
|
|
|
|
|
* siri/logger/erts-and-remote-log-olp:
[logger] Update bench marks to report percent instead of 0.XX
[logger] Minor fix in logger_olp_SUITE to avoid error in end_per_testcase
[logger] Remove info and reset functions from handler modules
[logger] Remove some unused variable warnings
[logger] Store proxy config in logger ets table
[logger] Log mode change and flushes in logger_proxy
[logger] Move out overload protection macros from logger_h_common.hrl
[logger] Allow logger_olp callbacks to return {stop,...}
Add logger_stress_SUITE to benchmarks spec
[logger] Add idle timer in logger_olp
[logger] Use persistent_term for storing proxy reference
Update preloaded
Use system_time instead of monotonic_time in log events
[logger] Add test for restart of logger proxy
[logger] Add API function for configuring logger proxy
[logger] Add logger_stress_SUITE
[logger] Add tests for logger_proxy
[logger] Overload protect logging from erts and remote nodes
[logger] Split overload protection functionality to own module
|
|
Add a link on the line number in cover output
OTP-15541
|
|
Fix encrypt_config_file and decrypt_config_file
OTP-15540
|
|
* raimo/stdlib/gen_statem-optimization/OTP-15452:
correct: Work around a compiler mis-optimization
|
|
This was not a compiler optimization that misfired, rather that
the code neede separate case clauses for when the timer was
running and not, so to not call erlang:cancel_timer/1 nor
maps:remove/2 in the case clause where only
a map update was needed before recursion.
See the comment in loop_timouts_cancel/13
|
|
|
|
* ingela/ssl/DES-EDE/OTP-15539:
ssl: Correct 3des_ede_cbc check
|
|
* raimo/stdlib/gen_statem-optimization/OTP-15452:
Write some more comments in the engine loop
Work around a compiler mis-optimization
Move out helper functions to engine loop
Handle #trans_opts{} fields in parse_actions() arguments
Optimize gen_statem engine loop
Modify generic_statem_transit benchmark
|
|
* bjorn/compiler/misc:
beam_ssa_type: Simplify is_singleton_type/1
beam_ssa_opt: Run ssa_opt_tuple_size early
beam_ssa_codegen: Remove forgotten and unreachable clause
beam_ssa_opt: Run the type optimization pass twice
beam_ssa_type: Eliminate redundant 'succeeded' instructions
|
|
* bjorn/compiler/fix-inlined-funs:
sys_core_inline: Kill *all* fun annotations when inlining
|
|
* bjorn/compiler/beam_validator/ERL-832:
Make the beam_validator smarter (again)
|
|
* hasse/fix_calls_to_io_format:
ssh: Adjust some calls to io:format()
xmerl: Adjust a call to io:format()
sasl: Adjust some calls to io:format()
common_test: Adjust some calls to io:format()
|
|
adds cprof to list of tools
|
|
|
|
Running ssa_opt_tuple_size early will give more opportunities
for optimizations.
|
|
fd682dd3b1dc corrected label generation for 'or', but forgot to
remove the old incorrect clause (that can no longer be reached).
|
|
The code will be significantly improved by running the
type optimization pass twice.
The ssa_opt_misc pass can be eliminated because everything it does
is also done by the type optimization pass.
|
|
Needed because of the optimizations in 48f20bd589fa69.
https://bugs.erlang.org/browse/ERL-832
|
|
|
|
The common subexpression optimization accidentally caused
unnecessary spilling of live registers. A rather ugly
workaround can be found in loop_timeouts_cancel.
|
|
sys_core_inline didn't kill fun annotations in variables,
which could lead to duplicated wrapper functions for funs.
That was basically harmless because the duplicated functions
were eventually discarded.
|
|
The beam_ssa_type pass would leave redundant 'succeeded' instructions,
and depend on the live optimization pass to remove them.
Update beam_ssa_type to remove redundant 'succeeded' instructions.
This will not improve the generated code, but will improve compilation
times since it eliminates instructions and variables.
|
|
|
|
|
|
|
|
|
|
This strategy improves speed benchmarks and reduces the
amount of garbage that is produced.
More:
* Handle transition options as loop function arguments instead
of in a record.
* Shorten the state_enter loop and make it more efficient
by making it more part of the loop and less a loop restart.
* Simplify the internal timeout action handling somewhat.
|
|
Could cause ssl to claim to support 3des_ede_cbc when cryptolib does not
|
|
Do the optimizations of bs_put* instructions in beam_ssa_opt
and remove the beam_bs pass. This can lead to a slight improvement
of compilation times.
|
|
There are two instructions that take string operands:
{bs_put_string,Fail,NumberOfBytes,{string,String}}
{bs_match_string,Fail,Register,NumberOfBits,{string,String}}
In the canonical BEAM code that is passed to beam_asm, string String
is currently represented as a list. (The string in bs_match_string is
a bitstring before the beam_z compiler pass.) That is wasteful,
because there will be unnecessary conversions between lists and
binaries.
Change the representation of String to be a binary.
Furthermore, bs_put_string is an optimization of a bs_put_binary
instruction with a literal binary operand. Currently, the
bs_put_string instruction is introduced in beam_bs. Delay the
introduction of bs_put_string to the beam_z pass. That will simplify
optimizations and allow us to do the optimization currently done
in beam_bs in a SSA pass in a future commit.
|
|
The optimization can be applied in a few more places if done
before ssa_opt_bsm_shortcut (for example, in unicode:cbv/2).
|
|
Also remove some compiler warnings from test suite.
|
