diff options
Diffstat (limited to 'erts/doc/src/notes.xml')
| -rw-r--r-- | erts/doc/src/notes.xml | 619 |
1 files changed, 617 insertions, 2 deletions
diff --git a/erts/doc/src/notes.xml b/erts/doc/src/notes.xml index 5866ee704a..a71d1eddf5 100644 --- a/erts/doc/src/notes.xml +++ b/erts/doc/src/notes.xml @@ -31,6 +31,334 @@ </header> <p>This document describes the changes made to the ERTS application.</p> +<section><title>Erts 10.1.1</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + A bug where the socket option 'pktoptions' caused a read + of uninitialized memory has been fixed. Would cause + malfunction on FreeBSD.</p> + <p> + Own Id: OTP-14297 Aux Id: OTP-15141 </p> + </item> + <item> + <p>Fixed a memory leak on errors when reading files.</p> + <p> + Own Id: OTP-15318</p> + </item> + <item> + <p>File access through UNC paths works again on Windows. + This regression was introduced in OTP 21.</p> + <p> + Own Id: OTP-15333 Aux Id: ERL-737 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.1</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fix the seq_trace token to not be cleared when a process + receives messages sent by erts. Some examples of when + this could happen is all port BIFs, i.e. + <c>open_port</c>, <c>port_command</c> etc etc.</p> + <p> + Fix so that messages sent by nifs can be traced using + normal and <c>seq_trace</c> tracing.</p> + <p> + Own Id: OTP-15038 Aux Id: ERL-602 </p> + </item> + <item> + <p> + Fixed specs and documentation for <c>process_info</c> + item <c>monitored_by</c> to include port identifiers and + nif resources as possible types.</p> + <p> + Own Id: OTP-15180 Aux Id: ERL-648 </p> + </item> + <item> + <p> + Fix bug in generation of erl_crash.dump, which could + cause VM to crash.</p> + <p> + Bug exist since erts-9.2 (OTP-20.2).</p> + <p> + Own Id: OTP-15181</p> + </item> + <item> + <p> + Fix bug where ctrl-break or ctrl-c would not trigger the + break mode properly on Windows. This bug was introduced + in erts-10.0 (OTP-21).</p> + <p> + Own Id: OTP-15205</p> + </item> + <item> + <p> + Fix a performance bug for reception of UDP packages, + where a memory buffer would be reallocated when it should + not have been.</p> + <p> + Introduce a limit on the maximum automatic increase of + the UDP user-space buffer to the theoretical max of the + network PATH, i.e. 65535.</p> + <p> + Own Id: OTP-15206</p> + </item> + <item> + <p> + Fix alignment of erts allocator state internally in erts. + With the improper alignment the emulator would refuse to + start when compiled with clang on 32-bit systems.</p> + <p> + Own Id: OTP-15208 Aux Id: PR-1897 ERL-677 </p> + </item> + <item> + <p> + Fix bug where too many concurrent calls to + <c>erlang:open_port({spawn,"cmd"},...)</c> would result + in the emulator terminating with the reason "Failed to + write to erl_child_setup: ". After this fix the + <c>open_port</c> call will throw an <c>emfile</c> + exception instead.</p> + <p> + Own Id: OTP-15210</p> + </item> + <item> + <p> + Upgraded the ERTS internal PCRE library from version 8.41 + to version 8.42. See <url + href="http://pcre.org/original/changelog.txt">http://pcre.org/original/changelog.txt</url> + for information about changes made to PCRE. This library + implements major parts of the <seealso + marker="stdlib:re"><c>re</c></seealso> regular + expressions module.</p> + <p> + Own Id: OTP-15217</p> + </item> + <item> + <p> + Fix <c>open_port({fd,X,Y}, ...)</c> to release the file + descriptors from the pollset when closing the port. + Without this fix the same file descriptor number could + not be reused when doing multiple open_port and + port_close sequences.</p> + <p> + Own Id: OTP-15236 Aux Id: ERL-692 </p> + </item> + <item> + <p> + Fixed bug in <c>float_to_list/2</c> and + <c>float_to_binary/2</c> with options + <c>[{decimals,0},compact]</c> causing totally wrong + results. Bug exists since OTP-21.0.</p> + <p> + Own Id: OTP-15276 Aux Id: PR-1920 </p> + </item> + <item> + <p> + Fixed bug in <c>erlang:memory</c> causing <c>ets</c> to + report too much. This small false memory leak (16 bytes + each time) can only happen when a specific race condition + occurs between scheduler threads on a table with option + <c>write_concurrency</c>.</p> + <p> + Own Id: OTP-15278</p> + </item> + <item> + <p> + Minor <c>configure</c> test fixes</p> + <p> + Own Id: OTP-15282</p> + </item> + <item> + <p> + Improved robustness of distribution connection setup. In + OTP-21.0 a truly asynchronous connection setup was + introduced. This is further improvement on that work to + make the emulator more robust and also be able to recover + in cases when involved Erlang processes misbehave.</p> + <p> + Own Id: OTP-15297 Aux Id: OTP-15279, OTP-15280 </p> + </item> + </list> + </section> + + + <section><title>Improvements and New Features</title> + <list> + <item> + <p> + The socket options <c>recvtos</c>, <c>recvttl</c>, + <c>recvtclass</c> and <c>pktoptions</c> have been + implemented in the socket modules. See the documentation + for the <c>gen_tcp</c>, <c>gen_udp</c> and <c>inet</c> + modules. Note that support for these in the runtime + system is platform dependent. Especially for + <c>pktoptions</c> which is very Linux specific and + obsoleted by the RFCs that defined it.</p> + <p> + Own Id: OTP-15145 Aux Id: ERIERL-187 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.0.8</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + As of ERTS version 10.0 (OTP 21.0) the + <c>erl_child_setup</c> program, which creates port + programs, ignores <c>TERM</c> signals. This setting was + unintentionally inherited by port programs. Handling of + <c>TERM</c> signals in port programs has now been + restored to the default behavior. That is, terminate the + process.</p> + <p> + Own Id: OTP-15289 Aux Id: ERIERL-235, OTP-14943, ERL-576 </p> + </item> + <item> + <p> + The fix made for OTP-15279 in erts-10.07 (OTP-21.0.8) was + not complete. It could cause a new connection attempt to + be incorrectly aborted in certain cases. This fix will + amend that flaw.</p> + <p> + Own Id: OTP-15296 Aux Id: OTP-15279, ERIERL-226 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.0.7</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + A process could get stuck in an infinite rescheduling + loop between normal and dirty schedulers. This bug was + introduced in ERTS version 10.0.</p> + <p> + Thanks to Maxim Fedorov for finding and fixing this + issue.</p> + <p> + Own Id: OTP-15275 Aux Id: PR-1943 </p> + </item> + <item> + <p> + Garbage collection of a distribution entry could cause an + emulator crash if <c>net_kernel</c> had not brought + previous connection attempts on it down properly.</p> + <p> + Own Id: OTP-15279 Aux Id: ERIERL-226 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.0.6</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + A race between termination of a process and resume of the + same process via <c>erlang:resume_process/1</c> could + cause the VM to crash. This bug was introduced in erts + version 10.0 (OTP 21.0).</p> + <p> + Own Id: OTP-15237</p> + </item> + <item> + <p> + When tracing on <c>running</c>, <c>in</c> trace events + could be lost when a process was rescheduled between a + dirty and a normal scheduler.</p> + <p> + Own Id: OTP-15269 Aux Id: ERL-713 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.0.5</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fixed a bug which caused an emulator crash when + <c>enif_send()</c> was called by a NIF that executed on a + dirty scheduler. The bug was either triggered when the + NIF called <c>enif_send()</c> without a message + environment, or when the process executing the NIF was + <c>send</c> traced.</p> + <p> + Own Id: OTP-15223</p> + </item> + <item> + <p> + Fixed a bug causing some Erlang references to be + inconsistently ordered. This could for example cause + failure to look up certain elements with references as + keys in search data structures. This bug was introduced + in R13B02.</p> + <p> + Thanks to Simon Cornish for finding the bug and supplying + a fix.</p> + <p> + *** POTENTIAL INCOMPATIBILITY ***</p> + <p> + Own Id: OTP-15225</p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 10.0.4</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p>Fixed a bug that prevented the <c>noshell</c> option + from working correctly on Mac OS X and BSD.</p> + <p> + Own Id: OTP-15169</p> + </item> + <item> + <p>Fixed a crash when matching directly against a literal + map using a single key that had been saved on the + stack.</p> + <p> + Own Id: OTP-15184</p> + </item> + <item> + <p>Fix node crash when passing a bad time option to + <c>file:read_file_info/2</c>.</p> + <p> + Own Id: OTP-15196</p> + </item> + </list> + </section> + +</section> + <section><title>Erts 10.0.3</title> <section><title>Fixed Bugs and Malfunctions</title> @@ -1027,6 +1355,108 @@ </section> +<section><title>Erts 9.3.3.5</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + ERTS internal trees of monitor structures could get into + an inconsistent state. This could cause <c>'DOWN'</c> + messages not to be delivered when they should, as well as + delivery of <c>'DOWN'</c> messages that should not be + delivered.</p> + <p> + This bug was introduced in ERTS version 9.0 (OTP 20.0) + and was fixed in ERTS version 10.0 (OTP 21.0) due to a + rewrite of the monitor code. That is, this bug only exist + in the OTP 20 release.</p> + <p> + Own Id: OTP-15399 Aux Id: ERL-751, ERIERL-262, OTP-14205 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 9.3.3.4</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fixed bug in <c>ets:select_replace</c> when called with a + fully bound key could cause a following call to + <c>ets:next</c> or <c>ets:prev</c> to crash the emulator + or return invalid result.</p> + <p> + Own Id: OTP-15346</p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 9.3.3.3</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fixed a bug which caused an emulator crash when + <c>enif_send()</c> was called by a NIF that executed on a + dirty scheduler. The bug was either triggered when the + NIF called <c>enif_send()</c> without a message + environment, or when the process executing the NIF was + <c>send</c> traced.</p> + <p> + Own Id: OTP-15223</p> + </item> + <item> + <p> + Fixed a bug causing some Erlang references to be + inconsistently ordered. This could for example cause + failure to look up certain elements with references as + keys in search data structures. This bug was introduced + in R13B02.</p> + <p> + Thanks to Simon Cornish for finding the bug and supplying + a fix.</p> + <p> + *** POTENTIAL INCOMPATIBILITY ***</p> + <p> + Own Id: OTP-15225</p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 9.3.3.2</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p>Fixed a race condition in the inet driver that could + cause receive to hang when the emulator was compiled with + gcc 8.</p> + <p> + Own Id: OTP-15158 Aux Id: ERL-654 </p> + </item> + <item> + <p> + Fix bug in generation of erl_crash.dump, which could + cause VM to crash.</p> + <p> + Bug exist since erts-9.2 (OTP-20.2).</p> + <p> + Own Id: OTP-15181</p> + </item> + </list> + </section> + +</section> + <section><title>Erts 9.3.3.1</title> <section><title>Fixed Bugs and Malfunctions</title> @@ -2469,6 +2899,79 @@ </section> +<section><title>Erts 8.3.5.6</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fixed small memory leak that could occur when sending to + a terminating port.</p> + <p> + Own Id: OTP-14609 Aux Id: ERIERL-238 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 8.3.5.5</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p>Fixed a race condition in the inet driver that could + cause receive to hang when the emulator was compiled with + gcc 8.</p> + <p> + Own Id: OTP-15158 Aux Id: ERL-654 </p> + </item> + <item> + <p> + Fixed a bug causing some Erlang references to be + inconsistently ordered. This could for example cause + failure to look up certain elements with references as + keys in search data structures. This bug was introduced + in R13B02.</p> + <p> + Thanks to Simon Cornish for finding the bug and supplying + a fix.</p> + <p> + *** POTENTIAL INCOMPATIBILITY ***</p> + <p> + Own Id: OTP-15225</p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 8.3.5.4</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p>Fixed a bug in file closure on Unix; close(2) was + retried on EINTR which could cause a different (recently + opened) file to be closed as well.</p> + <p> + Own Id: OTP-14775</p> + </item> + <item> + <p> + A race-condition when tearing down a connection with + active node monitors could cause the runtime system to + crash.</p> + <p> + This bug was introduced in ERTS version 8.0 (OTP 19.0).</p> + <p> + Own Id: OTP-14781 Aux Id: OTP-13047 </p> + </item> + </list> + </section> + +</section> + <section><title>Erts 8.3.5.3</title> <section><title>Fixed Bugs and Malfunctions</title> @@ -4296,6 +4799,22 @@ </section> +<section><title>Erts 7.3.1.5</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + Fixed small memory leak that could occur when sending to + a terminating port.</p> + <p> + Own Id: OTP-14609 Aux Id: ERIERL-238 </p> + </item> + </list> + </section> + +</section> + <section><title>Erts 7.3.1.4</title> <section><title>Fixed Bugs and Malfunctions</title> <list> @@ -5788,6 +6307,103 @@ </section> +<section><title>Erts 6.4.1.7</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + A process communicating with a port via one of the + <c>erlang:port_*</c> BIFs could potentially end up in an + inconsistent state if the port terminated during the + communication. When this occurred the process could later + block in a <c>receive</c> even though it had messages + matching in its message queue.</p> + <p> + This bug was introduced in erts version 5.10 (OTP R16A).</p> + <p> + Own Id: OTP-13424 Aux Id: OTP-10336 </p> + </item> + <item> + <p> + Calls to <c>erl_drv_send_term()</c> or + <c>erl_drv_output_term()</c> from a non-scheduler thread + while the corresponding port was invalid caused the + emulator to enter an inconsistent state which eventually + caused an emulator crash.</p> + <p> + Own Id: OTP-13866</p> + </item> + <item> + <p>Driver and NIF operations accessing processes or ports + could cause an emulator crash when used from + non-scheduler threads. Those operations are:</p> <list> + <item><c>erl_drv_send_term()</c></item> + <item><c>driver_send_term()</c></item> + <item><c>erl_drv_output_term()</c></item> + <item><c>driver_output_term()</c></item> + <item><c>enif_send()</c></item> + <item><c>enif_port_command()</c></item> </list> + <p> + Own Id: OTP-13869</p> + </item> + <item> + <p> + Fix bug in <c>binary_to_term</c> for binaries created by + <c>term_to_binary </c> with option <c>compressed</c>. The + bug can cause <c>badarg</c> exception for a valid binary + when Erlang VM is linked against a <c>zlib</c> library of + version 1.2.9 or newer. Bug exists since OTP 17.0.</p> + <p> + Own Id: OTP-14159 Aux Id: ERL-340 </p> + </item> + <item> + <p> + Fixed bug in operator <c>bxor</c> causing erroneuos + result when one operand is a big <em>negative</em> + integer with the lowest <c>N*W</c> bits as zero and the + other operand not larger than <c>N*W</c> bits. <c>N</c> + is an integer of 1 or larger and <c>W</c> is 32 or 64 + depending on word size.</p> + <p> + Own Id: OTP-14514</p> + </item> + <item> + <p> + Fixed bug in <c>binary_to_term</c> and + <c>binary_to_atom</c> that could cause VM crash. + Typically happens when the last character of an UTF8 + string is in the range 128 to 255, but truncated to only + one byte. Bug exists in <c>binary_to_term</c> since ERTS + version 5.10.2 (OTP_R16B01) and <c>binary_to_atom</c> + since ERTS version 9.0 (OTP-20.0).</p> + <p> + Own Id: OTP-14590 Aux Id: ERL-474 </p> + </item> + </list> + </section> + +</section> + +<section><title>Erts 6.4.1.6</title> + + <section><title>Fixed Bugs and Malfunctions</title> + <list> + <item> + <p> + When calling <c>garbage_collect/[1,2]</c> or + <c>check_process_code/[2,3]</c> from a process with a + higher priority than the priority of the process operated + on, the run queues could end up in an inconsistent state. + This bug has now been fixed.</p> + <p> + Own Id: OTP-13298 Aux Id: OTP-11388 </p> + </item> + </list> + </section> + +</section> + <section><title>Erts 6.4.1.5</title> <section><title>Fixed Bugs and Malfunctions</title> @@ -9879,7 +10495,7 @@ you use erlang:halt/2 with an integer first argument and an option list containing {flush,false} as the second argument. Note that now is flushing not dependant of the - exit code, and you can not only flush async threads + exit code, and you cannot only flush async threads operations which we deemed as a strange behaviour anyway. </p> <p>Also, erlang:halt/1,2 has gotten a new feature: If the @@ -18614,4 +19230,3 @@ </section> </section> </chapter> - |