| Age | Commit message (Collapse) | Author |
|---|
|
|
|
This also hides the module behind ?PRIM_FILE to make testing new
implementations less painful.
|
|
This improves the latency of file operations as dirty schedulers
are a bit more eager to run jobs than async threads, and use a
single global queue rather than per-thread queues, eliminating the
risk of a job stalling behind a long-running job on the same thread
while other async threads sit idle.
There's no such thing as a free lunch though; the lowered latency
comes at the cost of increased busy-waiting which may have an
adverse effect on some applications. This behavior can be tweaked
with the +sbwt flag, but unfortunately it affects all types of
schedulers and not just dirty ones. We plan to add type-specific
flags at a later stage.
sendfile has been moved to inet_drv to lessen the effect of a nasty
race; the cooperation between inet_drv and efile has never been
airtight and the socket dying at the wrong time (Regardless of
reason) could result in fd aliasing. Moving it to the inet driver
makes it impossible to trigger this by closing the socket in the
middle of a sendfile operation, while still allowing it to be
aborted -- something that can't be done if it stays in the file
driver.
The race still occurs if the controlling process dies in the short
window between dispatching the sendfile operation and the dup(2)
call in the driver, but it's much less likely to happen now.
A proper fix is in the works.
--
Notable functional differences:
* The use_threads option for file:sendfile/5 no longer has any
effect.
* The file-specific DTrace probes have been removed. The same
effect can be achieved with normal tracing together with the
nif__entry/nif__return probes to track scheduling.
--
OTP-14256
|
|
|
|
* maint:
Fix type create_option() in mnesia
|
|
Fix type create_option() in mnesia
|
|
* bjorn/compiler/coverage:
v3_codegen: Remove check of operand for bs_context_to_binary
beam_jump: Eliminate a repeated clause
beam_asm: No longer allow iolists as contents in chunk/2
beam_utils_SUITE: Cover more lines in beam_utils
guard_SUITE: Add a test case to cover beam_dead:turn_op/1
Recognize 'nil' as a literal in beam_utils:bif_to_test/3
Cover more code in beam_bsm:btb_opt_1/3
Add test to cover a line in v3_kernel:opt_single_valued/3
|
|
* maint:
Avoid falling measurements testcases on slow machines
stdlib: string optimize special case for ASCII
stdlib: Minor unicode_util opts
|
|
* dgud/stdlib/optimize-string-lists:
Avoid falling measurements testcases on slow machines
stdlib: string optimize special case for ASCII
stdlib: Minor unicode_util opts
OTP-14670
|
|
|
|
|
|
Avoid unicode_util module call for ASCII strings
|
|
Exit early for Latin-1
|
|
|
|
* ingela/ssl/ERL-521/OTP-14794:
ssl: Make sure all possible data is delivered
|
|
Conflicts:
lib/ssl/src/ssl_connection.erl
|
|
* ingela/ssl/cleaner-code:
ssl: Add gracefullness to dtls code
ssl: Remove old softupgrade code
ssl: Use genstamtem properly
|
|
The bs_context_to_binary instruction only allows a register operand.
v3_codegen has a test to ensure that the operand is a register.
That test is no longer necessary. There used to be a possibility
that optimizations in sys_core_fold and the inliner could change
the operand for bs_context_to_binary to a binary literal. Since
09112806c15a81b that can no longer happen, because no more
optimizations are run after the introduction of the
bs_context_to_binary instruction.
|
|
|
|
* maint:
dialyzer: Extend the map implementation's handling of ?unit
dialyzer: Use string:find() instead of string:str()
|
|
* hasse/dialyzer/maps_anon_fun/OTP-14743:
dialyzer: Extend the map implementation's handling of ?unit
dialyzer: Use string:find() instead of string:str()
|
|
* maint:
public_key: Fix failing pkix_verify_hostname_subjAltName_IP TC
|
|
* hans/ssh/public_key/fix_bad_cert_merge:
public_key: Fix failing pkix_verify_hostname_subjAltName_IP TC
|
|
The create_option() type in mnesia lacks the user_properties field
which is used by mnesia in various places.
|
|
* maint:
Update HOWTO with wxWidgets build examples
https should be mapped to url and not seealso
Update kernel appup for OTP-20.2
|
|
* dgud/erl_docgen/https2url:
Update HOWTO with wxWidgets build examples
https should be mapped to url and not seealso
|
|
* siri/appups-20.2:
Update kernel appup for OTP-20.2
|
|
Was due to an unresolved merge conflict in a453532fe31eecd8c39b5035702ee20b402fd687
|
|
This clause seems to have been introduced in cac51274eb9a.
|
|
The clause that converted an iolist to a binary was never
executed.
Note that chunk/2 is called for all chunks in the
{extra_chunks,Chunks} option. This change will enforce that the
contents of each chunk must be a binary (as documented).
|
|
|
|
f9a323d10a9f5d added consistent operand order for equality
comparisons. As a result, beam_dead:turn_op/1 is no longer covered.
We must keep the uncovered lines in beam_dead to ensure that
beam_dead can handle BEAM assembly code from another source than
v3_codegen that might not follow the operand order convention.
The only way to cover the lines is to use BEAM assembly in
the test case.
|
|
beam_utils:bif_to_test/3 is supposed to never put a literal
operand as the first operand in is_eq_exact or is_ne_exact,
but 'nil' was not recognized as a literal.
|
|
|
|
The uncovered line was added in 6753bbcc3fdb0.
|
|
Place move S x0 instructions at the end of blocks
|
|
* hans/public_key/base64_fix/OTP-14788:
public_key: Fix bad calls to base64:*
|
|
Also make tls code a little more direct for easier uderstanding
|
|
This code is only relevant in version is was written in.
|
|
|
|
|
|
* ingela/dtls/merge-fix:
ssl: Fix incorrect merge conflict resolution
|
|
When handling merging of back ported
Counter measurements for Bleichenbacher attack
a line from DTLS was accidentally lost.
|
|
* maint:
Use base64 encoding in crash dumps
Correct parsing of sub binaries
Generalize passing of options for decoding
|
|
* bjorn/base64-in-dumps/OTP-14686:
Use base64 encoding in crash dumps
Correct parsing of sub binaries
Generalize passing of options for decoding
|
|
The loader has a lot of fused instructions that include move S x0.
Placing them at the end of blocks makes it possible to take advantage
of this optimization more frequently.
|
|
|
|
|
|
|
|
|
