aboutsummaryrefslogtreecommitdiffstats
path: root/lib/hipe/doc/src/notes.xml
diff options
context:
space:
mode:
Diffstat (limited to 'lib/hipe/doc/src/notes.xml')
-rw-r--r--lib/hipe/doc/src/notes.xml96
1 files changed, 95 insertions, 1 deletions
diff --git a/lib/hipe/doc/src/notes.xml b/lib/hipe/doc/src/notes.xml
index 8c9dbc0c18..5c06e5e558 100644
--- a/lib/hipe/doc/src/notes.xml
+++ b/lib/hipe/doc/src/notes.xml
@@ -4,7 +4,7 @@
<chapter>
<header>
<copyright>
- <year>2006</year><year>2010</year>
+ <year>2006</year><year>2011</year>
<holder>Ericsson AB. All Rights Reserved.</holder>
</copyright>
<legalnotice>
@@ -30,6 +30,100 @@
</header>
<p>This document describes the changes made to HiPE.</p>
+<section><title>Hipe 3.7.9</title>
+
+ <section><title>Fixed Bugs and Malfunctions</title>
+ <list>
+ <item>
+ <p>
+ Fix erroneous fail info of a hipe_bs_primop</p>
+ <p>
+ Own Id: OTP-9036</p>
+ </item>
+ <item>
+ <p>
+ The change fixes a bug in the translation of 'bs_add'
+ BEAM instruction to HiPE's Icode representation. When
+ these instructions appeared in a guard context the
+ previous translation was obviously buggy.</p>
+ <p>
+ Own Id: OTP-9044</p>
+ </item>
+ <item>
+ <p>
+ Sanitize the specs of the code module</p>
+ <p>
+ After the addition of unicode_binary() to the
+ file:filename() type, dialyzer started complaining about
+ erroneous or incomplete specs in some functions of the
+ 'code' module. The culprit was hard-coded information in
+ erl_bif_types for functions of this module, which were
+ not updated. Since these functions have proper specs
+ these days and code duplication (pun intended) is never a
+ good idea, their type information was removed from
+ erl_bif_types.</p>
+ <p>
+ While doing this, some erroneous comments were fixed in
+ the code module and also made sure that the code now runs
+ without dialyzer warnings even when the
+ -Wunmatched_returns option is used.</p>
+ <p>
+ Some cleanups were applied to erl_bif_types too.</p>
+ <p>
+ Own Id: OTP-9100</p>
+ </item>
+ <item>
+ <p>
+ Fix bug in the simplification of inexact comparisons</p>
+ <p>
+ On 31/1/2011 Paul Guyot reported a bug in the native code
+ compilation of inexact equality/inequality tests between
+ floats and integers. The relevant test was:</p>
+ <p>
+ f(X) -&gt; Y = X / 2, Y == 0.</p>
+ <p>
+ and hipe erroneously evaluated the calls f(0) and f(0.0)
+ to 'false'.</p>
+ <p>
+ The culprit was in the simplification code of the Icode
+ range analysis which used an erroneous test (lists:any/1
+ instead of lists:all/1).</p>
+ <p>
+ Own Id: OTP-9101</p>
+ </item>
+ <item>
+ <p>
+ Document exiting and garbage_collecting process statuses</p>
+ <p>
+ Own Id: OTP-9102</p>
+ </item>
+ <item>
+ <p>
+ Remove hipe constants pool</p>
+ <p>
+ Hipe constants used to be allocated within a single,
+ fixed-size pool for interaction with the garbage
+ collector. However, the garbage collector no longer
+ depends on constants being allocated within a single
+ pool, and the fixed size of the pool both meant
+ unnecessary allocations on most deployments and crashes
+ on deployments requiring more constants.</p>
+ <p>
+ The code was simplified to directly invoke erts_alloc.
+ Debugging and undocumented function
+ hipe_bifs:show_literals/0 was removed (it returned true
+ and output text to the console), and debugging and
+ undocumented function hipe_bifs:constants_size/0 was
+ rewritten with a global to count the size of allocated
+ constants.</p>
+ <p>
+ Own Id: OTP-9128</p>
+ </item>
+ </list>
+ </section>
+
+</section>
+
<section><title>Hipe 3.7.8.1</title>
<section><title>Fixed Bugs and Malfunctions</title>