| Age | Commit message (Collapse) | Author |
|---|
|
Conflicts:
lib/stdlib/src/gen_statem.erl
|
|
* raimo/stdlib/optimize-gen_statem:
Optimize plain call response time
Correct typo in design principles for gen_statem
|
|
* maint:
kernel: Correct contracts and a bug in group_history
stdlib: Correct contracts
dialyzer: Optimize handling of a lot of warnings
Conflicts:
lib/kernel/src/erl_boot_server.erl
|
|
|
|
* maint:
stdlib: Garbage the shell's evaluator process more often
|
|
|
|
A minor fix.
If the Erlang shell saves no results (history(0) or results(0)), the
evaluator process now garbage collects itself. This means that any
huge binaries created by the evaluation are reclaimed faster than
before.
|
|
* maint:
stdlib: Handle Unicode when formatting stacktraces
|
|
* hasse/stdlib/unicode_stacktrace/OTP-14847/ERL-553:
stdlib: Handle Unicode when formatting stacktraces
|
|
* maint:
stdlib: Correct a filelib test case
stdlib: Let filelib:find_source() search subdirs
|
|
* hasse/stdlib/find_src/OTP-14832/ERL-527:
stdlib: Correct a filelib test case
stdlib: Let filelib:find_source() search subdirs
|
|
See also ERL-553 and ERL-544 (commit c3ddb0f).
|
|
* hasse/stdlib/erl_eval_stacktrace/OTP-14826/PR-1540:
syntax_tools: Correct handling of stacktrace variable
stdlib: Add check of stacktrace variable to erl_eval
stdlib: Improve erl_eval's stacktraces
|
|
* maint:
dialyzer: Add a test of erl_tar:table/1,2
Fix false Dialyzer warnings for erl_tar:table/1
|
|
Some of the functions of the erl_eval module do not call the Erlang
code linter, so they need to explicitly check that the newly
introduced stacktrace variable is not bound.
|
|
The call "erlang:get_stacktrace()" is not handled explicitly. If there
are issues, they can probably be ignored since erlang:get_stacktrace/1
will be deprecated and removed.
|
|
'tar_entry()' values are only returned if we specify the 'verbose'
option when calling table/2, which table/1 doesn't do.
Now, it appears that Dialyzer as of OTP 20 is clever enough to realize
that the return type of table/1 must intersect with the return type of
table/2, and so it ignores the fact that table/1 says it returns
strings, and therefore its callers are expected to be dealing with
'tar_entry()' tuples, and never with strings.
This is obviously a mismatch between what the code does and what the
spec says is does, leading to false Dialyzer warnings on code that uses
table/1 (and, presumably, also table/2 when called without
the 'verbose' option.)
|
|
The Design Principles states that an application can have Erlang
source files one level below the "src" directory, and now
filelib:find_source() by default searches one level below "src".
The same applies to "esrc". That directory is only mentioned in
filename(3).
|
|
* hasse/stdlib/gen_server_bench:
stdlib: Optimize gen a little
stdlib: Introduce gen_server benchmark
|
|
|
|
Add -MMD option to erlc
OTP-14830
|
|
Add syntax in try/catch to retrieve the stacktrace directly
|
|
* hasse/stdlib/base64/OTP-14624:
stdlib: Add base64 benchmarks
stdlib: Do not check base64 input more than needed
stdlib: Minor optimization of base64
stdlib: Use binary_to_list in base64 when it is faster
stdlib: Optimize base64 functions
|
|
The compile option makedep_side_effect, erlc -MMD, instructs
the compiler to emit dependencies and continue to compile
as normal.
|
|
add hibernate_after to the gen options type spec
|
|
This commit adds a new syntax for retrieving the stacktrace
without calling erlang:get_stacktrace/0. That allow us to
deprecate erlang:get_stacktrace/0 and ultimately remove it.
The problem with erlang:get_stacktrace/0 is that it can keep huge
terms in a process for an indefinite time after an exception. The
stacktrace can be huge after a 'function_clause' exception or a failed
call to a BIF or operator, because the arguments for the call will be
included in the stacktrace. For example:
1> catch abs(lists:seq(1, 1000)).
{'EXIT',{badarg,[{erlang,abs,
[[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20|...]],
[]},
{erl_eval,do_apply,6,[{file,"erl_eval.erl"},{line,674}]},
{erl_eval,expr,5,[{file,"erl_eval.erl"},{line,431}]},
{shell,exprs,7,[{file,"shell.erl"},{line,687}]},
{shell,eval_exprs,7,[{file,"shell.erl"},{line,642}]},
{shell,eval_loop,3,[{file,"shell.erl"},{line,627}]}]}}
2> erlang:get_stacktrace().
[{erlang,abs,
[[1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,
23,24|...]],
[]},
{erl_eval,do_apply,6,[{file,"erl_eval.erl"},{line,674}]},
{erl_eval,expr,5,[{file,"erl_eval.erl"},{line,431}]},
{shell,exprs,7,[{file,"shell.erl"},{line,687}]},
{shell,eval_exprs,7,[{file,"shell.erl"},{line,642}]},
{shell,eval_loop,3,[{file,"shell.erl"},{line,627}]}]
3>
We can extend the syntax for clauses in try/catch to optionally bind
the stacktrace to a variable.
Here is an example using the current syntax:
try
Expr
catch C:E ->
Stk = erlang:get_stacktrace(),
.
.
.
In the new syntax, it would look like:
try
Expr
catch
C:E:Stk ->
.
.
.
Only a variable (not a pattern) is allowed in the stacktrace position,
to discourage matching of the stacktrace. (Matching would also be
expensive, because the raw format of the stacktrace would have to be
converted to the cooked form before matching.)
Note that:
try
Expr
catch E ->
.
.
.
is a shorthand for:
try
Expr
catch throw:E ->
.
.
.
If the stacktrace is to be retrieved for a throw, the 'throw:'
prefix must be explicitly included:
try
Expr
catch throw:E:Stk ->
.
.
.
|
|
In the grammar file for the Erlang language, patterns are parsed
as expressions. erl_lint will then weed out expressions that are
not legal patterns.
The rule sharing causes problems if we were to introduce new
syntax, for example a pattern followed by a ':'. There would
be a shift/reduce conflict, and a pattern followed by a ':'
would be parsed as a remote call.
Introduce a new pat_expr rule to express exactly the subset
of expressions that is allowed in pattern.
Note: For the moment, we must allow full expressions in
case clauses to keep 'merl' working.
|
|
* maint:
Avoid falling measurements testcases on slow machines
stdlib: string optimize special case for ASCII
stdlib: Minor unicode_util opts
|
|
Avoid unicode_util module call for ASCII strings
|
|
|
|
Often the decode functions return a function_clause error, but not
always, and the errors have not been consistent between modifications
of the base64 module.
Now the errors are returned as they happen--no attempt to make them
look nice is done. The alternative, to ensure that, for example,
{badarg, Culprit} is always returned upon bad input, was deemed
pointless.
|
|
Using /bits instead of /binary is faster when constructing binaries.
|
|
It is sometimes faster to use binary_to_list on input, at least when
not considering time for garbage collections. The three functions all
return a list, and the temporary list created by binary_to_list should
be reclaimed very fast. The implementation from before is thus kept.
|
|
A few test cases with zeroes are added. They were not handled correctly
before.
The access of DECODE_MAP is moved into the inlined function b64d, for
symmetry.
The function b64e is also inlined. The speed-up is small, but
measurable.
Note: encode(List), decode(List) and mime_decode(List) no longer call
list_to_binary. This can break code that calls the functions with
I/O-lists as input.
|
|
Extend uri_string module (compose_query and dissect_query)
|
|
* lukas/stdlib/maps_iterators/OTP-14012:
erts: Limit size of first iterator for hashmaps
Update primary bootstrap
Update preloaded modules
erts: Remove erts_internal:maps_to_list/2
stdlib: Make io_lib and io_lib_pretty use maps iterator
erts: Implement batching maps:iterator
erts: Implement maps path iterator
erts: Implement map iterator using a stack
stdlib: Introduce maps iterator API
Conflicts:
bootstrap/lib/stdlib/ebin/io_lib.beam
bootstrap/lib/stdlib/ebin/io_lib_pretty.beam
erts/emulator/beam/bif.tab
erts/preloaded/ebin/erlang.beam
erts/preloaded/ebin/erts_internal.beam
erts/preloaded/ebin/zlib.beam
|
|
|
|
This iterator implementation fetches multiple elements to
iterate over in one call to erts_internal:maps_next instead
of one at a time. This means that the memory usage will go
up for the iterator as we are buffering elements, but the
usage is still bounded.
In this implementation the max memory usage is 1000 words.
Using this approach makes the iterator as fast as using
maps:to_list, so maps:iterator/2 has been removed.
|
|
Refactor supervisor, and store children in a map instead of a list
OTP-14586
|
|
Implement functions for handling form-urlencoded
query strings based on the HTML5 specification.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
compose_query/{1,2} and dissect_query/1 removed as the implemented
specification (HTML 2.0) is old. They will be re-implemented based
on HTML5.
|
|
|
|
|
|
|
