From: Joseph Jones <joeyi5216@...>
Date: 2015-12-17T21:05:40-07:00
Subject: [ruby-core:72238] [Ruby trunk - Bug #11790] [Open] [PATCH] ANSI alias rules fix

--56738614_19495cff_16c
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

Joseph Jones liked your message with Boxer. On December 8, 2015 at 09:32:=
36 MST, zarko=40ca.ibm.com wrote:Issue =2311790 has been reported by Zark=
o Todorovski.----------------------------------------Bug =2311790: =5BPAT=
CH=5D ANSI alias rules fix https://bugs.ruby-lang.org/issues/11790* Autho=
r: Zarko Todorovski* Status: Open* Priority: Normal* Assignee: * ruby -v:=
 * Backport: 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: UNKNOWN------------------=
----------------------line 70 in file regparse.h:*=23define SET=5FNTYPE(n=
ode, ntype) (node)->u.base.type =3D (ntype)*needs to be changed to confor=
m with ANSI alias rulesLine 70 as it is can lead to errors in compiling i=
n regparse.c*line 1143: node =3D (Node* )=46reeNodeList;**line 1144: =46r=
eeNodeList =3D =46reeNodeList->next;**line 1581: (node)->u.base.type =3D =
(0);* where during compiling line 1581 can be moved ahead of line 1144 si=
nce under ANSI aliasing rules, the compiler determines that the statement=
 of =22(node)->u.base.type =3D (0);=22 does not affect the value of =46re=
eNodeList. Proposed change in patch is in file regparse.sh:from: *=23defi=
ne SET=5FNTYPE(node, ntype) (node)->u.base.type =3D (ntype)*to:*=23define=
 SET=5FNTYPE(node, ntype) =7Bint value =3D ntype; memcpy(&((node)->u.base=
.type), &value, sizeof((node)->u.base.type));=7D*---=46iles--------------=
------------------alias=5Fansi=5Frule.patch (717 Bytes)-- https://bugs.ru=
by-lang.org/     
--56738614_19495cff_16c
Content-Type: text/html; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

<html><body><div>Joseph Jones liked your message with <a href=3D=22http:/=
/bxr.io/PBIqo=22>Boxer</a>.</div><br/><br/><div><div class=3D=22quote=22>=
On December 8, 2015 at 09:32:36 MST, zarko=40ca.ibm.com wrote:<br/><block=
quote type=3D=22cite=22 style=3D=22border-left-style:solid;border-width:1=
px;margin-left:0px;padding-left:10px;=22>Issue =2311790 has been reported=
 by Zarko  Todorovski.<br /><br />---------------------------------------=
-<br />Bug =2311790: =5BPATCH=5D ANSI alias rules fix <br />https://bugs.=
ruby-lang.org/issues/11790<br /><br />* Author: Zarko  Todorovski<br />* =
Status: Open<br />* Priority: Normal<br />* Assignee: <br />* ruby -v: <b=
r />* Backport: 2.0.0: UNKNOWN, 2.1: UNKNOWN, 2.2: UNKNOWN<br />---------=
-------------------------------<br />line 70 in file regparse.h:<br />*=23=
define SET=5FNTYPE(node, ntype)   (node)->u.base.type =3D (ntype)*<br /><=
br />needs to be changed to conform with ANSI alias rules<br /><br /><br =
/>Line 70 as it is can lead to errors in compiling in regparse.c<br /><br=
 />*line 1143: node =3D (Node* )=46reeNodeList;*<br />*line 1144: =46reeN=
odeList =3D =46reeNodeList->next;*<br />*line 1581: (node)->u.base.type =3D=
 (0);* <br /><br />where during compiling line 1581 can be moved ahead of=
 line 1144 since under ANSI aliasing rules, the compiler determines that =
the statement of =22(node)->u.base.type =3D (0);=22 does not affect the v=
alue of =46reeNodeList. <br /><br />Proposed change in patch is in file r=
egparse.sh:<br /><br />from: <br />*=23define SET=5FNTYPE(node, ntype)   =
(node)->u.base.type =3D (ntype)*<br />to:<br />*=23define SET=5FNTYPE(nod=
e, ntype)  =7Bint value =3D ntype; memcpy(&((node)->u.base.type), &value,=
 sizeof((node)->u.base.type));=7D*<br /><br />---=46iles-----------------=
---------------<br />alias=5Fansi=5Frule.patch (717 Bytes)<br /><br /><br=
 />-- <br />https://bugs.ruby-lang.org/<br /></blockquote></div></div></b=
ody></html>
--56738614_19495cff_16c--