| Age | Commit message (Collapse) | Author |
|---|
|
def_arg_reg[0] is used for storage of timeout instruction
when a 'receive after' is executed. When a process was
scheduled out inside prim_eval:'receive'/0 due to a function
call, def_arg_reg[0] was overwritten due to storage of live
registers.
prim_eval:'receive'/2 now calls arg_reg_alloc/0 which
bumps all reductions and then calls arg_reg_alloc/7
which will cause an allocation of a new arg_reg array
since def_arg_reg only can hold 6 values. This ensures
that the timeout instruction in def_arg_reg[0] used
for the timeout wont be overwritten.
|
|
* raimo/efile_drv-read_file-use-fstat/OTP-14184:
Use fstat if it exists in efile_openfile
|
|
* sverker/bin2term-zlib-bug/ERL-340/OTP-14159:
erts: Fix binary_to_term for compressed and zlib >= v1.2.9
|
|
|
|
Problem: z_stream was incorrectly copied with memcpy
which just happened to work with zlib < v1.2.9.
Solution: Avoid copying z_stream.
|
|
* sverker/make-export-fun-race/OTP-14144:
erts: Fix race bug between export fun creation and code loading
|
|
Symptom: SEGV crash on ARM in delete_code() -> export_list().
Could probably happen on other machines as well.
Problem: Staging export table was iterated in an unsafe way
while an entry was added for a new export fun.
Solution: Correct write order and some memory barriers.
|
|
|
|
* rickard/init-bug/OTP-13911:
Fix initialization dependencies between time, lock-check and lcnt
Conflicts:
erts/emulator/sys/unix/sys.c
|
|
again.
|
|
|
|
OTP-14051
* rickard/dirty-scheduling-fixes:
Update etp-commands for dirty schedulers
Fix scheduling of system tasks on processes executing dirty
Fix call time tracing with dirty schedulers
Fix send of exit signal to process executing dirty
Fix dirty scheduler process priority
Fix alloc-util hard-debug
Silence debug warning when no beam jump table is used with dirty schedulers
Conflicts:
erts/etc/unix/etp-commands.in
|
|
* rickard/nif-scheduling-fixes:
Fix check_process_code() when NifExport is in use - OTP-14048
Fix GC when NifExport is in use - OTP-14049
Fix saving of original arguments when rescheduling via NifExport - OTP-14050
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* sverker/erts/crash-dump-limit:
erts: Add env variable ERL_CRASH_DUMP_BYTES
erts: Add ErtsStrToSint64
erts: Refactor crash dumping with cbprintf
erts: Add cbprintf for Callback Printing
erts: Remove unused erl_crash_dump()
|
|
to limit crash dump size
|
|
for simplify string to 64-bit integer parsing.
|
|
Instead of passing around a file descriptor
use a function pointer to facilitate more advanced
backend write logic such as size limitation or compression.
|
|
|
|
* sverker/hipe-no-pie-amd64:
erts: Disable -fPIE for HiPE on x86_64
erts: Fix correct link flags for hipe_mkliterals
|
|
and no need for $(INCLUDED).
|
|
|
|
|
|
Do not generate a core when a crashdump is asked for.
Regression introduced in 56090db3ea417157a749bdd810fc61d117493f1f
|
|
* sverker/hipe-beam-stacktrace/ERL-205:
erts: Exclude random beam functions from hipe stacktrace
|
|
|
|
|
|
Avoid suspending fun caller not just if purge is already done
but also if purge of another module has started. Another purge
of the same module again cannot happen as making current to old
transition includes thread progress.
|
|
into maint
* lukas/erts/tracing/fix_dead_tracer_check/ERL-274/OTP-13928:
erts: Do tracer liveness check on current tracer
|
|
in Text part of error tuple, like
{error, {load, "Library load-call unsuccessful (606)}}
|
|
This fixes a fault introduced in 19.0 where an invalid
tracer would block setting of a new tracer on a process.
|
|
A process calling a fun from a module currently being soft-purged
could race with failure of the soft purge. When this happened the
call triggered loading of the module which erroneously would
load new code for the module in the case no new generation was
currently loaded.
|
|
|
|
|
|
* raimo/diffserv-socket-option/OTP-13582:
Tune 'tclass' semantics
Implement IPV6_TCLASS
|
|
* bjorn/erts/on_load/ERL-240/OTP-13893:
erts: Add nif_SUITE:t_on_load
erts: Improve nif_SUITE:upgrade test
Don't leak old code when loading a modules with an on_load function
|
|
|
|
|
|
|
|
|
|
to include upgrade from deleted (old) module instance
|
|
Normally, calling code:delete/1 before re-loading the code for a
module is unnecessary but causes no problem.
But there will be be problems if the new code has an on_load function.
Code with an on_load function will always be loaded as old code
to allowed it to be easily purged if the on_load function would fail.
If the on_load function succeeds, the old and current code will be
swapped.
So in the scenario where code:delete/1 has been called explicitly,
there is old code but no current code. Loading code with an
on_load function will cause the reference to the old code to be
overwritten. That will at best cause a memory leak, and at worst
an emulator crash (especially if NIFs are involved).
To avoid that situation, we will put the code with the on_load
function in a special, third slot in Module.
ERL-240
|
