annotate DOCS/gcc-2.96-3.0.html @ 2737:d0f34e15aa5f

vars are aligned by 8 now (even faster, yes they where not aligned correctly)
author michael
date Tue, 06 Nov 2001 13:45:57 +0000
parents f8352ff67ef7
children b2b3926a5e46
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
2440
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
1 <HTML>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
2 <BODY BGCOLOR=WHITE>
2441
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
3 <FONT face="Verdana, Arial, Helvetica, sans-serif" size=2>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
4
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
5 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
6 <B>Question:</B> What is GCC 2.96 ? I can't find it at GNU site.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
7 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
8
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
9 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
10 <B>Answer:</B> Read the <A HREF="http://gcc.gnu.org/gcc-2.96.html">official answer from GNU GCC team.</A>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
11 </P>
2440
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
12
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
13 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
14 <B>Question:</B> What is the problem with GCC 2.96 ? And with 3.x ?
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
15 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
16
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
17 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
18 <B>Answer:</B>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
19 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
20
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
21 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
22 And for the people, who periodically asks what are the exact problems with
2441
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
23 gcc 2.96, my answer: <I>we don't know exactly.</I>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
24 There were various problems and new problems / bugs come up periodically.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
25 It is <I>not a single bug/problem</I>. We just see various bugreports, mostly
2440
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
26 gcc internal bugs, compiler syntax errors in source or bad code compiled. They
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
27 all are solved using different version of gcc. I understand that gcc 2.96
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
28 has different default optimization flags and they conflicts with our inline
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
29 asm code, but we can't fix them, and we really don't want to fix them as they
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
30 work with other compilers or gcc versions, and the fix may cause speed loss.
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
31 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
32
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
33 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
34 I think that the gcc 2.96 should be fixed to be option-compatible with other
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
35 releases, but redhat guys refused to do it. If someone interested - ask
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
36 Eugene K., avifile author, he has a long mailing with them, because they had
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
37 the same problems with avifile. Finally he changed avifile source to
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
38 <I>workaround</I> gcc 2.96 bugs...
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
39 We simply has no interest and time to do it.
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
40 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
41
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
42 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
43 Ah, and about the pipe-in-comment bug: it wasn't really our bug.
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
44 I've talked one of gcc maintainers, and he told me that gcc 2.96 and 3.x
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
45 supports intel asm syntax, and it caused the pipe bug. But it was a bug,
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
46 because gcc silently, without any warning, ignored the whole asm block.
2441
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
47 <I>They</I> have fixed that, now it prints warning and doesn't skip the block.
2440
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
48 (at least he told me, i didn't checked)
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
49 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
50
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
51 <P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
52 Other gcc 3.x problems comes from broken libstdc++ or glibc header (std_*.h)
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
53 installation. They are not our fault. MPlayer compiles and works well with
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
54 gcc 3.x versions. <B>Only 2.96 is broken</B>, but it depends on many environment
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
55 elements, including gcc 2.96 release number, enabled mplayer features, etc.
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
56 <I>If it works for you using gcc 2.96, it doesn't mean it will work for everyone.</I>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
57 </P>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
58
2441
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
59 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
60 <B>Question:</B> No! You are wrong! Everything works with gcc 2.96 <I>but</I> MPlayer
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
61 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
62
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
63 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
64 <B>Answer:</B>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
65 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
66
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
67 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
68 No. You are wrong!
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
69 Several projects (mainly which source contains high optimized inline asm code)
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
70 had problems with gcc 2.96. For example: avifile, MESA / DRI, ffmpeg.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
71 But other projects already workarounded gcc bugs (changed code which
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
72 triggered compiler bugs) so they work for now.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
73 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
74
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
75 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
76 <B>Question:</B> No! You are wrong! Everything works with gcc 2.96 <I>including</I> MPlayer
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
77 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
78
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
79 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
80 <B>Answer:</B>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
81 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
82
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
83 Good. Be happy. But you must know, it depends on many environment
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
84 elements, including gcc 2.96 release number, enabled mplayer features, etc.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
85 <I>If it works for you using gcc 2.96, it doesn't mean it will work for everyone!</I>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
86 It only means that you are lucky, until you find a problem. But don't forget the
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
87 <B>No.1 rule of gcc 2.96 users: NEVER REPORT BUGS OR PROBLEMS IF YOU ARE USING GCC 2.96</B>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
88
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
89 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
90 <B>Question:</B> Ok. Understood. But I want to give it a try... how to compile with gcc 2.96?
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
91 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
92
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
93 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
94 <B>Answer:</B> Really? Are you sure? Ok. You know... here is it: ./configure --disable-gcc-checking
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
95 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
96
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
97 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
98 <B>Question:</B> No! I don't agree with you, because ...
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
99 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
100
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
101 <P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
102 <B>Answer:</B> It doesn't matter. Keep your commets for yourself. We're not interested in gcc 2.96 stories.
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
103 </P>
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
104
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
105
f8352ff67ef7 htmlize, added new things
arpi
parents: 2440
diff changeset
106 </FONT>
2440
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
107 </BODY>
3b792a1d0fc2 hmm. am I a doxwriter?
arpi
parents:
diff changeset
108 </HTML>