mirror/ioccc
1
0 Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge pull request #2705 from xexyl/clean-up-bugs

Browse Source
This commit is contained in:
Landon Curt Noll authored and GitHub committed 2024-10-19 09:13:33 -07:00
2 parents 4841532208 0ea1e00af7
commit 5abfa3fdcf
2 files changed
+320 -350
Download Patch File Download Diff File Expand all files Collapse all files

No files matched your search

bugs.html
+160 -173
View File
@@ -443,14 +443,13 @@ might seem to be fine but is actually not! We will note some where this is known
to happen (if it’s not yet fixed). This has actually already happened to us. to happen (if it’s not yet fixed). This has actually already happened to us.
Sometimes it is okay: some entries are known to not work but they can still have Sometimes it is okay: some entries are known to not work but they can still have
improvements (rarer situation but it’s happened). Some might have slightly improvements (rarer situation but it’s happened). Some might have slightly
different output but which is not a problem for instance a newline after output different output but which is not a problem for instance a newline after output.
in <a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a>’s changing it to This is not the most ideal but it is okay.</p>
<a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1994/ldb/ldb.c">1994/ldb</a> use <code>fgets()</code> from <code>gets()</code>.</p>
<p>Make <strong>ABSOLUTE CERTAIN</strong> that you read the index.html file <em>BEFORE</em> your changes <p>Make <strong>ABSOLUTE CERTAIN</strong> that you read the index.html file <em>BEFORE</em> your changes
as it’s important to see that the code is doing what it is supposed to. In the as it’s important to see that the code is doing what it is supposed to. In the
case that it’s not obvious (some index.html files do not even have commands to case that it’s not obvious (some index.html files do not even have commands to
try or even explain how to use it!) then you should either skip the entry OR ask try or even explain how to use it!) then you should either skip the entry OR ask
someone who will know e.g. us.</p> someone who will know for instance the Judges.</p>
<p>Make <strong>ABSOLUTE CERTAIN</strong> that you read the next section, the list of statuses <p>Make <strong>ABSOLUTE CERTAIN</strong> that you read the next section, the list of statuses
and the related information, BEFORE you submit a pull request!</p> and the related information, BEFORE you submit a pull request!</p>
<p><strong>BE VERY AWARE</strong> that sometimes fixing an entry for one platform will break it <p><strong>BE VERY AWARE</strong> that sometimes fixing an entry for one platform will break it
@@ -480,17 +479,17 @@ different warnings.</p>
sizes on different platforms. These tend to be required at the risk that sizes on different platforms. These tend to be required at the risk that
sometimes the entry will not work for certain platforms, some of which might or sometimes the entry will not work for certain platforms, some of which might or
might not be fixable; a good example where it was required to change and is okay might not be fixable; a good example where it was required to change and is okay
is when <a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed the is the fix of the segfault in macOS of
segfault in macOS of <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1989/paul/paul.c">1989/paul</a>: changing the <code>int *</code> to a <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1989/paul/paul.c">1989/paul</a>: changing the <code>int *</code> to a <code>long *</code>
<code>long *</code> was required and it works just as well with linux.</p> was required and it works just as well with linux.</p>
<p>But even if they are fixable (which will likely be hard to do) it’s almost <p>But even if they are fixable (which will likely be hard to do) it’s almost
certain that such code would be just as non-portable (importable ? :-) ).</p> certain that such code would be just as non-portable (importable ? :-) ).</p>
<p>In rare cases they are valid. For instance Cody noticed in linux the following <p>In rare cases they are valid. For instance in linux the following
warning in <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1985/applin/applin.c">1985/applin</a> with gcc:</p> warning in <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1985/applin/applin.c">1985/applin</a> with gcc:</p>
<pre><code> applin.c:1:78: warning: incompatible implicit declaration of built-in function &#39;execlp&#39; [-Wbuiltin-declaration-mismatch] <pre><code> applin.c:1:78: warning: incompatible implicit declaration of built-in function &#39;execlp&#39; [-Wbuiltin-declaration-mismatch]
1 | main(v,c)char**c;{for(v[c++]=strdup(&quot;Hello, world!\n&quot;);(!!c)[*c]&amp;&amp;(v--||--c&amp;&amp;execlp(*c,*c,c[!!c]+!!c,!c));**c=!c)write(!!*c,*c,!!**c);} 1 | main(v,c)char**c;{for(v[c++]=strdup(&quot;Hello, world!\n&quot;);(!!c)[*c]&amp;&amp;(v--||--c&amp;&amp;execlp(*c,*c,c[!!c]+!!c,!c));**c=!c)write(!!*c,*c,!!**c);}
| ^~~~~~</code></pre> | ^~~~~~</code></pre>
<p>and this reminded him of his fix to <p>and this is is kind of like the fix to
<a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/anonymous/anonymous.c">1984/anonymous</a>. In the case of <code>1984/anonymous</code> it <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/anonymous/anonymous.c">1984/anonymous</a>. In the case of <code>1984/anonymous</code> it
was a bit more involved but with <code>1985/applin</code> one need only add to the Makefile was a bit more involved but with <code>1985/applin</code> one need only add to the Makefile
<code>-include unistd.h</code>. So there are some cases where fixing warnings can fix a <code>-include unistd.h</code>. So there are some cases where fixing warnings can fix a
@@ -507,14 +506,14 @@ include you can update the Makefile <code>CINCLUDE</code> variable. For instance
needs <code>stdio.h</code> you could do <code>-include stdio.h</code>. Please leave a space after the needs <code>stdio.h</code> you could do <code>-include stdio.h</code>. Please leave a space after the
<code>=</code> in the Makefile. You may also have extra long lines if this seems useful to <code>=</code> in the Makefile. You may also have extra long lines if this seems useful to
make it a one-liner even if it kind of makes it longer than what the judges make it a one-liner even if it kind of makes it longer than what the judges
consider a one-liner. Thank you!</p> consider a one-liner, at least within reason. Thank you!</p>
<h3 id="on-layout-of-program-source">On layout of program source:</h3> <h3 id="on-layout-of-program-source">On layout of program source:</h3>
<p>If you make changes <em>PLEASE TRY AND KEEP THE SOURCE CODE LAYOUT AS CLOSE TO THE <p>If you make changes <em>PLEASE TRY AND KEEP THE SOURCE CODE LAYOUT AS CLOSE TO THE
ORIGINAL AS POSSIBLE</em>. This might not always be possible and if you have an ORIGINAL AS POSSIBLE</em>. This might not always be possible and if you have an
editor that does formatting it can cause problems. Sometimes formatters can even editor that does formatting it can cause problems. Sometimes formatters can even
break code! <a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> has break code! Some have experienced this many times with vim so he tends to
experienced this many times with vim so he tends to disable all format options disable all format options when formatting code. In vim you should be able to do
when formatting code. In vim you should be able to do that with:</p> that with:</p>
<pre><code> :set formatoptions=</code></pre> <pre><code> :set formatoptions=</code></pre>
<h2 id="status-known-bug">STATUS: known bug</h2> <h2 id="status-known-bug">STATUS: known bug</h2>
<p><strong>Please help us fix</strong>!</p> <p><strong>Please help us fix</strong>!</p>
@@ -564,7 +563,7 @@ much as possible given the nature of how it generates code: the entry itself
compiles but two versions of code it generates does not work with clang.</p> compiles but two versions of code it generates does not work with clang.</p>
<p>However some still might have problems of some kind or another and some might <p>However some still might have problems of some kind or another and some might
not yet be located as an additional defect of clang was noticed where <code>main()</code> not yet be located as an additional defect of clang was noticed where <code>main()</code>
is not allowed four args or one arg.</p> is not allowed four args or one arg (see below).</p>
<p>This is what the status means, however.</p> <p>This is what the status means, however.</p>
<p>Entries with this status have a problem in that the args to main() are not of a <p>Entries with this status have a problem in that the args to main() are not of a
specific type or there are too many due to this being allowed in earlier C. Some specific type or there are too many due to this being allowed in earlier C. Some
@@ -572,24 +571,24 @@ compilers like clang have a defect where they do not allow this and some
versions have an additional defect where they only allow 0, 2 or 3 args, the versions have an additional defect where they only allow 0, 2 or 3 args, the
latter of which affected <code>1989/westley</code>, so these entries do not work with latter of which affected <code>1989/westley</code>, so these entries do not work with
clang.</p> clang.</p>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> looked at the source <p>Looking at the source
code of clang and reported that there is no way to override the requirement of code of clang, it appears that there is no way to override the requirement of
arg types so these entries will fail to compile with clang. That’s why in some arg types so these entries will fail to compile with clang. That’s why in some
entries he fixed he did it by adding a function called <code>pain()</code>. :-) He did not entries the fix was done with a new function, often called <code>pain()</code> :-)</p>
check for the number of args as that was discovered later but he suspects it’s
the same issue.</p>
<p><strong>NOTE for macOS users</strong>: please be aware that <em>gcc</em> under macOS <strong>is actually <p><strong>NOTE for macOS users</strong>: please be aware that <em>gcc</em> under macOS <strong>is actually
clang</strong> despite the fact it might appear to be gcc: no symlink and both gcc and clang</strong> despite the fact it might appear to be gcc: no symlink and both gcc and
clang exist but the gcc is clang which you’ll see if you run <code>gcc --version</code>.</p> clang exist but the gcc is clang which you’ll see if you run <code>gcc --version</code>.</p>
<p>A tip and some fix methods from Cody: in the older days args to main() not given <p>A tip and some fix methods from <a href="authors.html#Cody_Boone_Ferguson">Cody Boone
a type were implicit ints but when they’re required to be <code>char **</code> this can Ferguson</a>: in the older days args to main()
cause a problem. In some cases Cody was able to use a <code>char *</code> inside <code>main()</code> not given a type were implicit ints but when they’re required to be <code>char **</code>
(see <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1989/tromp/tromp.c">1989/tromp/tromp.c</a> and this can cause a problem. In some cases it was possible to use a <code>char *</code> inside
<a href="1986/holloway/index.html">1986/holloway/holloway.c</a> for two examples though done slightly <code>main()</code> (see <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1989/tromp/tromp.c">1989/tromp/tromp.c</a> and
differently). In other cases he was able to dereference the pointers to be used <a href="1986/holloway/index.html">1986/holloway/holloway.c</a> for two examples though
like an int and other times a cast was necessary. He used other various done slightly differently). In other cases it was possible to dereference the
techniques to get them to compile. In some cases this introduced a problem but pointers to be used like an int and other times a cast was necessary. There are
typically if not always that problem exists with compilers that are less strict.</p> various techniques to get these to compile. In some cases this introduced a
problem but typically if not always that problem exists with compilers that are
less strict.</p>
<h2 id="status-main-has-only-one-arg">STATUS: main() has only one arg</h2> <h2 id="status-main-has-only-one-arg">STATUS: main() has only one arg</h2>
<p><strong>Please help modify so that main() as 2 or 3 args</strong>!</p> <p><strong>Please help modify so that main() as 2 or 3 args</strong>!</p>
<p>Because some versions of clang complain about the number of args of <code>main()</code> and <p>Because some versions of clang complain about the number of args of <code>main()</code> and
@@ -611,37 +610,35 @@ or if you can <code>-ggdb3</code> when compiling as that will help with debuggin
<h2 id="status-uses-gets">STATUS: uses gets()</h2> <h2 id="status-uses-gets">STATUS: uses gets()</h2>
<p><strong>Please help us change use of gets() to fgets()</strong>, if possible.</p> <p><strong>Please help us change use of gets() to fgets()</strong>, if possible.</p>
<p>Entries with this status use <code>gets()</code> which is unsafe because it has no limit on <p>Entries with this status use <code>gets()</code> which is unsafe because it has no limit on
the length of the string. <a href="authors.html#Cody_Boone_Ferguson">Cody Boone the length of the string. Many were fixed but not all.</p>
Ferguson</a> has fixed several of these and he’s <p>There’s a very good reason for this change even though it is bordering on tampering
working on others in between fixing other things.</p>
<p>There’s a very good reason for this even though it is bordering on tampering
entries: in modern systems one will get possibly alarming and certainly annoying entries: in modern systems one will get possibly alarming and certainly annoying
warnings during at least one of compiling, linking and execution. In macOS it’s warnings during at least one of compiling, linking and execution. In macOS it’s
most annoying as it happens during execution (also compilation) with the warning most annoying as it happens during execution (also compilation) with the warning
being interspersed with the output of the program, often causing confusing being interspersed with the output of the program, often causing confusing
output with the entry. A good example that was fixed is output with the entry. A good example that was fixed is
<a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/tbr/tbr.c">1990/tbr</a>.</p> <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/tbr/tbr.c">1990/tbr</a>.</p>
<p>Cody provided some important notes on this status with respect to fixing / <p>Some important notes on this status with respect to fixing / changing the
changing entries:</p> entries:</p>
<ol start="0" type="1"> <ol start="0" type="1">
<li><p>In some cases changing the entry to use <code>fgets()</code> will break the entry and in <li><p>In some cases changing the entry to use <code>fgets()</code> will break the entry and in
some cases cause it to crash. Where this is known we will document it. Some of some cases cause it to crash. Where this is known we will document it.</p></li>
these Cody is also off and on working on.</p></li>
<li><p>You <strong>MUST</strong> check the output before and after to make sure that it remains <li><p>You <strong>MUST</strong> check the output before and after to make sure that it remains
the same. Sometimes the output might not be immediately obviously wrong but is. the same. Sometimes the output might not be immediately obviously wrong but is.
We will document known examples as we come across them.</p></li> We will document known examples as we come across them.</p></li>
<li><p><code>gets()</code> does NOT STORE the <code>'\n'</code> but <code>fgets()</code> DOES! This is one of the <li><p><code>gets()</code> does NOT STORE the <code>'\n'</code> but <code>fgets()</code> DOES! This is one of the
reasons some of the entries break when changed to <code>fgets()</code>. Cody has fixed some reasons some of the entries break when changed to <code>fgets()</code>. Some of those have
of these but there are others he hasn’t had time to address or even discover been fixed but there are others that have not been.</p></li>
yet.</p></li>
<li><p>Just because you don’t see the string <code>gets</code> in the code does not mean it’s <li><p>Just because you don’t see the string <code>gets</code> in the code does not mean it’s
not used. It is after all the International <em>Obfuscated C Code</em> Contest! :-) not used. It is after all the International <em>Obfuscated C Code</em> Contest! :-)
Sometimes it will be in the Makefile and other times it will be obfuscated in Sometimes it will be in the Makefile and other times it will be obfuscated in
other ways. Compilers and linkers tend to warn about its use (and as noted in other ways; a good example of that that was fixed is
macOS it also happens at execution) and this is a good way to find entries that <a href="1991/dds/index.html">1991/dds</a>: in this case the string was encrypted.
use it even if it’s not visible in the code.</p></li> Compilers and linkers tend to warn about its use (and as noted in macOS it also
happens at execution) and this is a good way to find entries that use it even if
it’s not visible in the code.</p></li>
</ol> </ol>
<p><em>NOTE</em>: this status is <em>NOT necessarily mutually exclusive</em> with the <em>INABIAF</em> <p><strong>NOTE</strong>: this status is <em>NOT necessarily mutually exclusive</em> with the <em>INABIAF</em>
(it’s not a bug it’s a feature) status. The reason for this is due to warnings (it’s not a bug it’s a feature) status. The reason for this is due to warnings
during compiling, linking and/or runtime, sometimes causing confusing output (as during compiling, linking and/or runtime, sometimes causing confusing output (as
noted above).</p> noted above).</p>
@@ -683,7 +680,7 @@ version that works in modern systems) is very system specific and was before
system specific winning entries were discouraged. This is an all time personal system specific winning entries were discouraged. This is an all time personal
favourite of Landon Curt Noll. Run the alternate code to understand why this favourite of Landon Curt Noll. Run the alternate code to understand why this
might be (along with how strange the source code is :-) ).</p> might be (along with how strange the source code is :-) ).</p>
<p>An example where a crash is not a bug: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2019/endoh/prog.c">2019/endoh</a> is <p>An example where a crash is not a bug: <a href="2019/endoh/index.html">2019/endoh</a> is
supposed to crash. There are others that are also supposed to crash or that are supposed to crash. There are others that are also supposed to crash or that are
known to segfault but are considered features.</p> known to segfault but are considered features.</p>
<p>An important note is that if the index.html of the entry has a bug status that <p>An important note is that if the index.html of the entry has a bug status that
@@ -695,7 +692,7 @@ figure it out you’re invited to look at the git diffs, where there are some
(some were fixed earlier on but rolled back as both Cody and Landon individually (some were fixed earlier on but rolled back as both Cody and Landon individually
felt that the fix was tampering with the entry).</p> felt that the fix was tampering with the entry).</p>
<p><strong>NOTE</strong>: in the case of <code>gets()</code> we’ve fixed some to avoid the warning of the <p><strong>NOTE</strong>: in the case of <code>gets()</code> we’ve fixed some to avoid the warning of the
compiler, linker or even during runtime, depending on the system. In <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/tbr/tbr.c">some cases compiler, linker or even during runtime, depending on the system. In <a href="1990/tbr/index.html">some cases
like 1990/tbr</a> the fix actually prevents confusing output (though that like 1990/tbr</a> the fix actually prevents confusing output (though that
was not the only fix made in that entry).</p> was not the only fix made in that entry).</p>
<h3 id="exception-your-own-entries">Exception: your own entries</h3> <h3 id="exception-your-own-entries">Exception: your own entries</h3>
@@ -733,16 +730,17 @@ code. In particular you should see something like:</p>
<h3 id="status-inabiaf---please-do-not-fix-3">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-3">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-1984mullendermullender.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/mullender/mullender.c">1984/mullender/mullender.c</a></h3> <h3 id="source-code-1984mullendermullender.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/mullender/mullender.c">1984/mullender/mullender.c</a></h3>
<h3 id="information-1984mullenderindex.html">Information: <a href="1984/mullender/index.html">1984/mullender/index.html</a></h3> <h3 id="information-1984mullenderindex.html">Information: <a href="1984/mullender/index.html">1984/mullender/index.html</a></h3>
<p>Although there is an alt version and supplementary program added by Cody that <p>Although there is an alt version and supplementary program added, that
will work in modern systems, if you do not have a will work in modern systems, if you do not have a
<a href="https://en.wikipedia.org/wiki/VAX-11">VAX-11</a> or <a href="https://en.wikipedia.org/wiki/VAX-11">VAX-11</a> or
<a href="https://en.wikipedia.org/wiki/PDP-11">PDP-11</a> to run the original entry on it <a href="https://en.wikipedia.org/wiki/PDP-11">PDP-11</a> to run the original entry on it
will not work. See the index.html for details on the alternate versions.</p> will not work. See the index.html for details on the alternate versions.</p>
<p>Cody added and fixed the <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/mullender/gentab.c">gentab.c</a> which is from the <p>The file <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1984/mullender/gentab.c">gentab.c</a> which is from the
author’s (or one of them, Mullender) remarks found by Cody. Cody fixed this to author’s (or one of them, Mullender) remarks, was fixed for modern systems, at
compile and work (as best as he can determine: he has no VAX-11 or PDP-11 or least as best as can be determined: running the code on the binary itself does
emulator to test it) but running the code on the binary itself produces a produce a <code>short[]</code> that can compile in modern systems but it will not work.
<code>short[]</code> that can compile in modern systems though it’ll not work.</p> Before the fix it would fail in some cases and it is possible that in some cases
it might still fail.</p>
<hr style="width:10%;text-align:left;margin-left:0"> <hr style="width:10%;text-align:left;margin-left:0">
<div id="1985"> <div id="1985">
<h1 id="section-1">1985</h1> <h1 id="section-1">1985</h1>
@@ -823,8 +821,10 @@ reverse of other code (also wrt names). See the source file and the index.html
(in the author’s remarks) for more details.</p> (in the author’s remarks) for more details.</p>
<p>Fixing the (mis)feature is likely to be a very difficult challenge especially <p>Fixing the (mis)feature is likely to be a very difficult challenge especially
without breaking something else which is far more likely (see below in tips from without breaking something else which is far more likely (see below in tips from
Cody). You are welcome to try and fix it if you can!</p> Cody, who fixed it so the original would compile with <code>clang</code> and at least one
<h3 id="tips-from-cody">Tips from Cody:</h3> generated version would compile with <code>clang</code> too, not compromising the other
versions with <code>gcc</code>). You are welcome to try and fix it if you can!</p>
<h3 id="tips">Tips:</h3>
<p>The reason this is crashing is that the array <code>irk</code> is being accessed way out of <p>The reason this is crashing is that the array <code>irk</code> is being accessed way out of
bounds by the int <code>gnat</code>. For instance:</p> bounds by the int <code>gnat</code>. For instance:</p>
<pre><code> Program terminated with signal SIGSEGV, Segmentation fault. <pre><code> Program terminated with signal SIGSEGV, Segmentation fault.
@@ -836,12 +836,11 @@ bounds by the int <code>gnat</code>. For instance:</p>
<h3 id="magic-of-the-entry">Magic of the entry:</h3> <h3 id="magic-of-the-entry">Magic of the entry:</h3>
<p>The real trouble is that the code is generated and in a complex way or rather <p>The real trouble is that the code is generated and in a complex way or rather
ways.</p> ways.</p>
<p>I will explain this a little bit (though it might be more than a bit according <p>Trying to explain this might be tricky but here is an attempt. The program, when
to many :-) ) but that is all I can do for now. The program, when used without used without any args, will print out what it reads. If one arg is specified it
any args, will print out what it reads. If one arg is specified it will do a will do a ROT13 of it. If two args are specified it will print out the text
ROT13 of it. If two args are specified it will print out the text backwards backwards (<em>WITHOUT ROT13!</em>). If three args are specified it will do both
(<em>WITHOUT ROT13!</em>). If three args are specified it will do both <strong><em>reversal AND <strong><em>reversal AND ROT13</em></strong>.</p>
ROT13</em></strong>.</p>
<p>So <code>ver0</code> is the same as the main program; <code>ver1</code> is the ROT13 of the code (but <p>So <code>ver0</code> is the same as the main program; <code>ver1</code> is the ROT13 of the code (but
see below); <code>ver2</code> will reverse the code and <code>ver3</code> will both reverse and ROT13 see below); <code>ver2</code> will reverse the code and <code>ver3</code> will both reverse and ROT13
the code. But it’s trickier than that as one might expect.</p> the code. But it’s trickier than that as one might expect.</p>
@@ -879,19 +878,17 @@ just the reversal version, of course but ver1 is the ROT13 version.</p>
and then another version that has the reversal <em>and</em> ROT13 then you can see it’s and then another version that has the reversal <em>and</em> ROT13 then you can see it’s
nigh impossible to fix if not impossible.</p> nigh impossible to fix if not impossible.</p>
<p>The fix to get it to even compile with clang and then get the first two <p>The fix to get it to even compile with clang and then get the first two
(counting the original, <code>ver0</code>) to work was the hardest fix I made due to how (counting the original, <code>ver0</code>) to work was quite hard due to how the comments
the comments in reverse and ROT13 of it all works together, especially allowing in reverse and ROT13 of it all works together, especially allowing all versions
all versions to compile with gcc. The only reason that it doesn’t work to compile with gcc. The only reason that it doesn’t work completely with clang
completely with clang is the arg types of main() (see if you can figure out how is the arg types of main() (see if you can figure out how the change of <code>main()</code>
the change of <code>main()</code> in <code>westley.c</code> is not carried over to versions 2 and 3!); in <code>westley.c</code> is not carried over to versions 2 and 3!); originally none of
originally none of them compiled with clang.</p> them compiled with clang.</p>
<p>But it was possible to get the ROT13 code to compile with clang but the other <p>But it was possible to get the ROT13 code to compile with clang but the other
versions, reversed as well as ROT13 and reversed code, proved much more versions, reversed as well as ROT13 and reversed code, proved much more
problematic. I got it to compile (or maybe mostly compile) all versions with problematic. It is now possible to compile (or maybe mostly compile) all versions with
clang but I then carried it over to gcc in another system and it had syntax clang but on another system and it had syntax errors.</p>
errors.</p> <p>Enjoy! :-)</p>
<p>That’s about all I can say for how it works as other things have to be done too.
Enjoy! :-)</p>
<hr style="width:10%;text-align:left;margin-left:0"> <hr style="width:10%;text-align:left;margin-left:0">
<div id="1990"> <div id="1990">
<h1 id="section-6">1990</h1> <h1 id="section-6">1990</h1>
@@ -968,11 +965,10 @@ more time and resources to run as well. For instance:</p>
<h3 id="status-known-bug---please-help-us-fix-1">STATUS: known bug - please help us fix</h3> <h3 id="status-known-bug---please-help-us-fix-1">STATUS: known bug - please help us fix</h3>
<h3 id="source-code-1990jawjaw.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/jaw/jaw.c">1990/jaw/jaw.c</a></h3> <h3 id="source-code-1990jawjaw.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/jaw/jaw.c">1990/jaw/jaw.c</a></h3>
<h3 id="information-1990jawindex.html">Information: <a href="1990/jaw/index.html">1990/jaw/index.html</a></h3> <h3 id="information-1990jawindex.html">Information: <a href="1990/jaw/index.html">1990/jaw/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> <p>Some issues in this program were fixed and <a href="authors.html#Yusuke_Endoh">Yusuke
fixed some issues in this program and <a href="authors.html#Yusuke_Endoh">Yusuke Endoh</a> then provided the <code>btoa</code> script but it appears
Endoh</a> provided the <code>btoa</code> there is a bug in this entry. The judges wrote that to test the entry one can
script but it appears there is a bug in this entry. The judges wrote that to do:</p>
test the entry one can do:</p>
<pre><code> echo &quot;Quartz glyph jocks vend, fix, BMW.&quot; | compress | ./btoa | ./jaw</code></pre> <pre><code> echo &quot;Quartz glyph jocks vend, fix, BMW.&quot; | compress | ./btoa | ./jaw</code></pre>
<p>which should apply the identity transformation to a minimal holoalphabetic <p>which should apply the identity transformation to a minimal holoalphabetic
sentence.</p> sentence.</p>
@@ -1011,18 +1007,18 @@ defeat the purpose. This would ideally be fixed.</p>
<h3 id="status-inabiaf---please-do-not-fix-10">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-10">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-1990theoremtheorem.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/theorem/theorem.c">1990/theorem/theorem.c</a></h3> <h3 id="source-code-1990theoremtheorem.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/theorem/theorem.c">1990/theorem/theorem.c</a></h3>
<h3 id="information-1990theoremindex.html">Information: <a href="1990/theorem/index.html">1990/theorem/index.html</a></h3> <h3 id="information-1990theoremindex.html">Information: <a href="1990/theorem/index.html">1990/theorem/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed many bugs that <p>Many bugs that prevented this from working properly (including segfaults) were
prevented this from working properly (including segfaults) but one thing to note fixed but one thing to note is that if you pass two zeroes to <code>theorem_bkp</code> or
is that if you pass two zeroes to <code>theorem_bkp</code> or <code>fibonacci</code> the program <code>fibonacci</code> the program will enter an infinite loop, printing 0 over and over
will enter an infinite loop, printing 0 over and over again; another condition again; another condition where this occurred was fixed but this one should not
where this occurred was fixed but this one should not be fixed. Thank you.</p> be fixed.</p>
<div id="1990_westley"> <div id="1990_westley">
<h2 id="westley-1">1990/westley</h2> <h2 id="westley-1">1990/westley</h2>
</div> </div>
<h3 id="status-inabiaf---please-do-not-fix-11">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-11">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-1990westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/westley/westley.c">1990/westley/westley.c</a></h3> <h3 id="source-code-1990westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1990/westley/westley.c">1990/westley/westley.c</a></h3>
<h3 id="information-1990westleyindex.html">Information: <a href="1990/westley/index.html">1990/westley/index.html</a></h3> <h3 id="information-1990westleyindex.html">Information: <a href="1990/westley/index.html">1990/westley/index.html</a></h3>
<p>Although Cody fixed this to not enter an infinite loop if the arg (converted to <p>Although this was fixed to not enter an infinite loop if the arg (converted to
a number) is &lt; 0 the lack of an arg check at all was kept in to make it like the a number) is &lt; 0 the lack of an arg check at all was kept in to make it like the
original. The reason for the &lt; 0 check is it floods the screen (yes this is original. The reason for the &lt; 0 check is it floods the screen (yes this is
indeed inconsistent with some other entries but it is not worth the time spent, indeed inconsistent with some other entries but it is not worth the time spent,
@@ -1067,9 +1063,8 @@ when you’re cheating it ends up winning! Can you figure that out as well?</p>
<h3 id="source-code-1992adrianadrian.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/adrian/adrian.c">1992/adrian/adrian.c</a></h3> <h3 id="source-code-1992adrianadrian.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/adrian/adrian.c">1992/adrian/adrian.c</a></h3>
<h3 id="information-1992adrianindex.html">Information: <a href="1992/adrian/index.html">1992/adrian/index.html</a></h3> <h3 id="information-1992adrianindex.html">Information: <a href="1992/adrian/index.html">1992/adrian/index.html</a></h3>
<p>The author stated that if the file cannot be opened then it will print a system <p>The author stated that if the file cannot be opened then it will print a system
error but this is not the case unless it’s showing a fault. However this was error but this is not the case unless it’s showing a fault. This was fixed as
instead fixed by Cody as part of another problem that was detected so that the part of another problem that was detected.</p>
program just exits 1 if the file cannot be opened.</p>
<p>Also, if <em>some</em> of the programs are not given an arg they will very likely crash <p>Also, if <em>some</em> of the programs are not given an arg they will very likely crash
or do something funny. For instance <code>adsleep</code> and <code>adbasename</code>. This is very or do something funny. For instance <code>adsleep</code> and <code>adbasename</code>. This is very
simple to fix but the author explicitly noted that this will segfault and it is simple to fix but the author explicitly noted that this will segfault and it is
@@ -1103,7 +1098,7 @@ in <code>albert.orig.c</code> that has been fixed in <code>albert.c</code>,</p>
Both albert and albert.orig loop without printing anything, although Both albert and albert.orig loop without printing anything, although
the first factor is 101 and is usually found in an instant.</p> the first factor is 101 and is usually found in an instant.</p>
</blockquote> </blockquote>
<p>A quick debugging session from Cody suggests that the problem with this value <p>A quick debugging session suggests that the problem with this value
is that:</p> is that:</p>
<pre><code> if ( ppp-&gt;qq!=48 ) return;</code></pre> <pre><code> if ( ppp-&gt;qq!=48 ) return;</code></pre>
<p>is never reached because the value is never 48 (it is in other cases but not for <p>is never reached because the value is never 48 (it is in other cases but not for
@@ -1170,13 +1165,13 @@ click the mouse it takes it back towards the centre.</p>
<h3 id="status-doesnt-work-with-some-compilers---please-provide-alternative-code-or-fix-for-more-compilers">STATUS: doesn’t work with some compilers - please provide alternative code or fix for more compilers</h3> <h3 id="status-doesnt-work-with-some-compilers---please-provide-alternative-code-or-fix-for-more-compilers">STATUS: doesn’t work with some compilers - please provide alternative code or fix for more compilers</h3>
<h3 id="source-code-1992lushlush.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/lush/lush.c">1992/lush/lush.c</a></h3> <h3 id="source-code-1992lushlush.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/lush/lush.c">1992/lush/lush.c</a></h3>
<h3 id="information-1992lushindex.html">Information: <a href="1992/lush/index.html">1992/lush/index.html</a></h3> <h3 id="information-1992lushindex.html">Information: <a href="1992/lush/index.html">1992/lush/index.html</a></h3>
<p>We used a patch from <a href="authors.html#Yusuke_Endoh">Yusuke Endoh</a> to get this to <p>This was fixed but it only properly works with gcc; it was also changed to use
work but it only works with gcc. Cody removed the warnings of <code>gets()</code> and <code>fgets()</code> instead of <code>gets()</code> and a script was provided to run the entry
provided a script to run the entry properly.</p> properly. Nonetheless it will only work with gcc.</p>
<p>Unfortunately due to the way the entry works and the fact that other compilers <p>Unfortunately due to the way the entry works and the fact that other compilers
like clang have different warnings and errors this simply does not work with like clang have different warnings and errors this simply does not work with
them.</p> them.</p>
<p>Some tips from Cody:</p> <p>Some tips:</p>
<p>This entry relies on specific compiler warnings. With gcc it will look something <p>This entry relies on specific compiler warnings. With gcc it will look something
like:</p> like:</p>
<pre><code> [...] <pre><code> [...]
@@ -1244,7 +1239,7 @@ it yourself through ctrl-c or killing it in some other fashion.</p>
<h3 id="status-inabiaf---please-do-not-fix-17">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-17">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-1992westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/westley/westley.c">1992/westley/westley.c</a></h3> <h3 id="source-code-1992westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1992/westley/westley.c">1992/westley/westley.c</a></h3>
<h3 id="information-1992westleyindex.html">Information: <a href="1992/westley/index.html">1992/westley/index.html</a></h3> <h3 id="information-1992westleyindex.html">Information: <a href="1992/westley/index.html">1992/westley/index.html</a></h3>
<p>Cody improved the usability of this program by making it so that as long as the <p>The usability of this program was improved by making it so that as long as the
terminal columns is &gt;= 80 it will display properly, rather than having to wrap terminal columns is &gt;= 80 it will display properly, rather than having to wrap
at 80 columns. However due to the nature of the program if the terminal is &lt; 80 at 80 columns. However due to the nature of the program if the terminal is &lt; 80
in column width it will not display right. To see the number of columns in your in column width it will not display right. To see the number of columns in your
@@ -1359,12 +1354,12 @@ there.</p>
<h3 id="status-inabiaf---please-do-not-fix-24">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-24">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-1994ldbldb.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1994/ldb/ldb.c">1994/ldb/ldb.c</a></h3> <h3 id="source-code-1994ldbldb.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1994/ldb/ldb.c">1994/ldb/ldb.c</a></h3>
<h3 id="information-1994ldbindex.html">Information: <a href="1994/ldb/index.html">1994/ldb/index.html</a></h3> <h3 id="information-1994ldbindex.html">Information: <a href="1994/ldb/index.html">1994/ldb/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed this to compile <p>This was fixed to compile with modern systems but the entry also used <code>gets(3)</code>
with modern systems but the entry also used <code>gets(3)</code> which in some systems would which in some systems would print out a warning along with the output of the
print out a warning along with the output of the program. Naturally it could program. Naturally it could also overflow long lines, of which there are many,
also overflow long lines, of which there are many, though that is more though that is more considered a feature, not a bug, as it was documented by the
considered a feature, not a bug, as it was documented by the author.</p> author.</p>
<p>Cody changed it to <code>fgets(3)</code> to prevent the display problem but this introduces <p>It was changed to use <code>fgets(3)</code> to prevent the display problem but this introduces
another problem namely that newlines can be printed if the line length &lt; 231.</p> another problem namely that newlines can be printed if the line length &lt; 231.</p>
<p>This seems like a worthy compromise to not have messed up output and although it <p>This seems like a worthy compromise to not have messed up output and although it
would be ideal for it to never print a newline (unless that’s the line itself) would be ideal for it to never print a newline (unless that’s the line itself)
@@ -1383,30 +1378,28 @@ for more information on the change to <code>fgets(3)</code>.</p>
even when it works. See the author’s remarks in the index.html for details. even when it works. See the author’s remarks in the index.html for details.
Increasing the buffer size and having it, when fed its own source code, Increasing the buffer size and having it, when fed its own source code,
generate code that will compile with the same buffer size is difficult: it just generate code that will compile with the same buffer size is difficult: it just
uses the original size. Cody explains this further down.</p> uses the original size. This is explained later down.</p>
<p>Getting it to use <code>fgets()</code> is even harder as when fed its own code it will <p>Getting it to use <code>fgets()</code> is even harder as when fed its own code it will
generate code that cannot even compile let alone use <code>fgets(3)</code>.</p> generate code that cannot even compile let alone use <code>fgets(3)</code>.</p>
<p>In both cases, changing the buffer size and changing it to use <code>fgets(3)</code>, will <p>In both cases, changing the buffer size and changing it to use <code>fgets(3)</code>, will
cause compilation errors without other adjustments.</p> cause compilation errors without other adjustments.</p>
<p>See the below magic for details. Along with the author’s remarks in the <p>See the below magic for details. Along with the author’s remarks in the
index.html file it might prove possible to get it to use <code>fgets(3)</code>. Cody, index.html file it might prove possible to get it to use <code>fgets(3)</code>.</p>
writing this on 02 November 2023, only just noticed the author’s remarks and <p>Here is the magic of how this entry works, to help anyone who wishes to work on
will later on look at this if nobody takes up the challenge. More important work this.</p>
like getting to a place that the next contest can run must be done first.</p> <p>You might wish to run the commands:</p>
<p>Cody explains the magic of how this entry works, which will be necessary if this <pre><code> diff schnitzi.alt.c schnitzi.alt2.c
entry is to be fixed, below.</p>
<p>You might wish to run the command:</p> diff schnitzi.alt2.c schnitzi.alt.c</code></pre>
<pre><code> make diff_orig_prog</code></pre> <p>to see how the buffer size and how the <code>gets(3)</code> was changed to use <code>fgets()</code>,
<p>to see what had to change for the buffer size, when looking at the below. in the two alternate versions. Notice that the <code>fgets(3)</code> version can compile it
Furthermore you will want to look at:</p> just can’t <strong>generate</strong> compilable code (let alone using <code>fgets(3)</code>) when fed
<pre><code> diff schnitzi.alt.c schnitzi.alt2.c</code></pre> itself (to reiterate, just changing the call to <code>fgets(3)</code> does not mean it can
<p>as the <code>fgets(3)</code> version can compile it just can’t generate compilable code (let compile and there’s a further problem in that <code>fgets(3)</code> retains the newline
alone using <code>fgets(3)</code>) when fed itself (to reiterate, just changing the call to whereas <code>gets(3)</code> does not). Nevertheless looking at these commands will be of
<code>fgets(3)</code> does not mean it can compile and there’s a further problem in that help to understand how it works, in case one wishes to try and fix it.</p>
<code>fgets(3)</code> retains the newline whereas <code>gets(3)</code> does not). Nevertheless looking
at these commands will be of help to understand how it works.</p>
<p>For the alternate versions the other functionality is unaffected.</p> <p>For the alternate versions the other functionality is unaffected.</p>
<h3 id="the-magic-of-1994schnitzi-and-how-it-flips-text">The magic of <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1994/schnitzi/schnitzi.c">1994/schnitzi</a> and how it flips text</h3> <h3 id="the-magic-of-1994schnitzi-and-how-it-flips-text">The magic of <a href="1994/schnitzi/index.html">1994/schnitzi</a> and how it flips text</h3>
<p>The problem is getting the generated code to use <code>fgets()</code> (once it even <p>The problem is getting the generated code to use <code>fgets()</code> (once it even
compiles which was easy to do) and also have the updated buffer size be the compiles which was easy to do) and also have the updated buffer size be the
same (which was easy to do too at least in the original version). The generated same (which was easy to do too at least in the original version). The generated
@@ -1450,9 +1443,9 @@ changing the buffer size, see below).</p>
<p>Getting this entry to use <code>fgets(3)</code> is easy but the problem is you’re supposed <p>Getting this entry to use <code>fgets(3)</code> is easy but the problem is you’re supposed
to be able to feed the source to the program and the output of that will be to be able to feed the source to the program and the output of that will be
compilable. It however will create compiler errors. So it’s not just changing compilable. It however will create compiler errors. So it’s not just changing
the code to get it to use <code>fgets(3)</code>! I (Cody) already did this before I noticed the code to get it to use <code>fgets(3)</code>! This was done but
the other part which is why I rolled that part back. I did increase the buffer the other part is why that part was rolled back. The buffer size was increased
size but that only works on the original source.</p> but that only works on the original source.</p>
<p>To say just how sensitive this entry is: even a space, lack of space or a <p>To say just how sensitive this entry is: even a space, lack of space or a
character, wrong character, lack of character or character in the wrong place character, wrong character, lack of character or character in the wrong place
can cause a compilation error! Make sure that the output of:</p> can cause a compilation error! Make sure that the output of:</p>
@@ -1468,7 +1461,7 @@ compiled!</p>
<p>This program will likely crash if the source code file (by the name of the file <p>This program will likely crash if the source code file (by the name of the file
that’s compiled) cannot be opened in the directory it is run from.</p> that’s compiled) cannot be opened in the directory it is run from.</p>
<h3 id="important-reminder-and-a-note-about-the--1-value-check-for-getc">Important reminder and a note about the <code>-1</code> value check for <code>getc()</code>:</h3> <h3 id="important-reminder-and-a-note-about-the--1-value-check-for-getc">Important reminder and a note about the <code>-1</code> value check for <code>getc()</code>:</h3>
<p>Cody fixed the code to not use <code>-1</code> for the return value of <code>getc()</code>; this is <p>This code was fixed to not use <code>-1</code> for the return value of <code>getc()</code>; this is
important because <code>EOF</code> is <strong>NOT</strong> guaranteed to be <code>-1</code> but rather any negative important because <code>EOF</code> is <strong>NOT</strong> guaranteed to be <code>-1</code> but rather any negative
value. On systems where <code>EOF != -1</code> the program would enter an infinite loop value. On systems where <code>EOF != -1</code> the program would enter an infinite loop
until the program crashed, by chance reads a <code>-1</code> or was killed (it is for this until the program crashed, by chance reads a <code>-1</code> or was killed (it is for this
@@ -1515,8 +1508,7 @@ as documented, are considered features.</p>
<h3 id="source-code-1995cduacdua.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1995/cdua/cdua.c">1995/cdua/cdua.c</a></h3> <h3 id="source-code-1995cduacdua.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/1995/cdua/cdua.c">1995/cdua/cdua.c</a></h3>
<h3 id="information-1995cduaindex.html">Information: <a href="1995/cdua/index.html">1995/cdua/index.html</a></h3> <h3 id="information-1995cduaindex.html">Information: <a href="1995/cdua/index.html">1995/cdua/index.html</a></h3>
<p>This did not originally compile under macOS and after it did compile under <p>This did not originally compile under macOS and after it did compile under
macOS, it crashed. <a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> macOS, it crashed. These problems were fixed.</p>
fixed these problems.</p>
<p>It should be noted however that there is a condition where the program will <p>It should be noted however that there is a condition where the program will
prompt you to press return again. This was thought to be a bug but looking at prompt you to press return again. This was thought to be a bug but looking at
the code it can clearly be seen that if <code>g - a</code> is 0 then the message is the code it can clearly be seen that if <code>g - a</code> is 0 then the message is
@@ -1754,7 +1746,7 @@ with things but we don’t know.</p>
<pre><code> $ echo poot | ./dlowe <pre><code> $ echo poot | ./dlowe
poot was here!</code></pre> poot was here!</code></pre>
<p>Can you help fix these problems?</p> <p>Can you help fix these problems?</p>
<p>Cody provided the C code that creates an unformatted perl script that will just <p>Here is some C code that creates an unformatted perl script that will just
print ‘unimplemented’ for operators but it seems to work in some other ways (see print ‘unimplemented’ for operators but it seems to work in some other ways (see
below for how to go further than unimplemented) which is the exact perl from the below for how to go further than unimplemented) which is the exact perl from the
C program except that it has the <code>#!/usr/bin/env perl</code> and <code>use warnings;</code> as C program except that it has the <code>#!/usr/bin/env perl</code> and <code>use warnings;</code> as
@@ -1834,18 +1826,17 @@ For instance inputting <code>D</code> followed by <code>XX</code>:</p>
<p>That means then that in linux despite the subroutines seemingly not existing it <p>That means then that in linux despite the subroutines seemingly not existing it
works in the code, at least in some cases.</p> works in the code, at least in some cases.</p>
<p>As far as the unimplemented error goes: the author stated that it means there is <p>As far as the unimplemented error goes: the author stated that it means there is
invalid input. But as can be seen by the input I (Cody) gave the input is not invalid input. But as can be seen by the input given, the input is not
actually incorrect. It is also curious to note that it ends up printing what’s actually incorrect. It is also curious to note that it ends up printing what’s
on the stack at that point:</p> on the stack at that point:</p>
<pre><code> $ echo 5 5 + p | ./dlowe.pl <pre><code> $ echo 5 5 + p | ./dlowe.pl
unimplemented unimplemented
5</code></pre> 5</code></pre>
<p>which might (?) suggest that the <code>+</code> operator is unimplemented. Unfortunately it <p>which might (?) suggest that the <code>+</code> operator is unimplemented.</p>
has been many years since I have used perl and I was never a guru either.</p>
<h3 id="status-inabiaf---please-do-not-fix-32">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-32">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<p>The author states that in perl &lt; 5.6.0 there is a bug with a core dump in what <p>The author states that in perl &lt; 5.6.0 there is a bug with a core dump in what
they said is in <code>Perl_sv_upgrade</code>. As this is documented it is not considered a they said is in <code>Perl_sv_upgrade</code>. As this is documented it is not considered a
bug to be fixed. For the curious this will crash in macOS. Cody notes that the bug to be fixed. For the curious this will crash in macOS. It was noted that the
code that crashes is:</p> code that crashes is:</p>
<pre><code> perl_eval_sv <pre><code> perl_eval_sv
(newSVpv(&quot;_:$_= &lt;&gt; ; defined or exit; @ARGV&quot; (newSVpv(&quot;_:$_= &lt;&gt; ; defined or exit; @ARGV&quot;
@@ -1919,16 +1910,17 @@ fire</a>! :-) ).</p>
<h3 id="status-inabiaf---please-do-not-fix-35">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-35">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-2001anonymousanonymous.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.c">2001/anonymous/anonymous.c</a></h3> <h3 id="source-code-2001anonymousanonymous.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.c">2001/anonymous/anonymous.c</a></h3>
<h3 id="information-2001anonymousindex.html">Information: <a href="2001/anonymous/index.html">2001/anonymous/index.html</a></h3> <h3 id="information-2001anonymousindex.html">Information: <a href="2001/anonymous/index.html">2001/anonymous/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed this so that it <p>This was fixed so that it doesn’t segfault and then also fixed the functionality
doesn’t segfault and then also fixed the functionality of it (but see below).</p> of it (but see below).</p>
<p>Cody also fixed the <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.ten.c">anonymous.ten.c</a> program <p>The <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.ten.c">anonymous.ten.c</a> program was
(it also segfaulted) but notes that it <strong>MUST</strong> be compiled as a <strong>32-bit ELF also fixed (it also segfaulted) but note that it <strong>MUST</strong> be compiled as a <strong>32-bit ELF
binary</strong> so it will NOT work in macOS (for example); if you cannot use <code>-m32</code> or binary</strong> so it will NOT work in macOS (for example); if you cannot use <code>-m32</code> or
even if you can but it cannot compile as an ELF binary (not elf binary :-) ) even if you can but it cannot compile as an ELF binary (not elf binary :-) )
(but see below) then the <code>anonymous</code> program itself will very likely crash if (but see below) then the <code>anonymous</code> program itself will very likely crash if
you run it on the compiled <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.ten.c">anonymous.ten.c</a> but you run it on the compiled
if nothing else it will not modify the target executable (this part of the fix <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/anonymous/anonymous.ten.c">anonymous.ten.c</a>; if nothing else
at least should be correct).</p> it will not modify the target executable (this part of the fix at least should
be correct).</p>
<p>Note also that if you don’t specify a file or you specify a non-32-bit ELF file <p>Note also that if you don’t specify a file or you specify a non-32-bit ELF file
this program will very likely crash or do something strange like slaughter the this program will very likely crash or do something strange like slaughter the
elves of Imladris :-(</p> elves of Imladris :-(</p>
@@ -1940,36 +1932,35 @@ elves of Imladris :-(</p>
<h3 id="information-2001bellardindex.html">Information: <a href="2001/bellard/index.html">2001/bellard/index.html</a></h3> <h3 id="information-2001bellardindex.html">Information: <a href="2001/bellard/index.html">2001/bellard/index.html</a></h3>
<p>The author stated that it only works with i386 linux so the fact it doesn’t work <p>The author stated that it only works with i386 linux so the fact it doesn’t work
in modern systems is considered a feature and not a bug.</p> in modern systems is considered a feature and not a bug.</p>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed an initial <p>An initial segfault was fixed (trying to open the file) and the <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/bellard/bellard.otccex.c">supplementary
segfault (trying to open the file) and he also fixed the <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/bellard/bellard.otccex.c">supplementary program program bellard.otccex.c</a> was also
bellard.otccex.c</a> but this still crashes in fixed but this still crashes in systems not i386 linux which is what the author
systems not i386 linux which is what the author stated.</p> stated.</p>
<p>The author said that they compiled it with <a href="https://ftp.gnu.org/gnu/gcc/gcc-2.95.2/gcc-everything-2.95.2.tar.gz">gcc version <p>The author said that they compiled it with <a href="https://ftp.gnu.org/gnu/gcc/gcc-2.95.2/gcc-everything-2.95.2.tar.gz">gcc version
2.95.2</a>. We 2.95.2</a>. We
don’t know if a certain gcc version is necessary but it might be helpful to don’t know if a certain gcc version is necessary but it might be helpful to
download and compile that version to test it - or it might not.</p> download and compile that version to test it - or it might not.</p>
<p>I (Cody) have no i386 system to test this but perhaps this is why I can’t get it <p>Yusuke was able to get this to work with <code>-m32</code> but only in an emulator. As
to work. Yusuke was able to get this to work with <code>-m32</code> but only in an such, it might be that the changes made should be rolled back but at least now
emulator.</p> the supplementary program will work in modern systems.</p>
<p>On the author’s <a href="https://bellard.org/otcc/">web page for this program</a> it is <p>On the author’s <a href="https://bellard.org/otcc/">web page for this program</a> it is
explicitly stated that it requires i386 linux.</p> explicitly stated that it requires i386 linux.</p>
<p>There I found what should be a more portable version which is included as <p>There it was found what should be a more portable version which is included as
<a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/bellard/otccelf.c">otccelf.c</a> (after adding some <code>#include</code>s and the <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/bellard/otccelf.c">otccelf.c</a> (after adding some <code>#include</code>s
modification by Yusuke noted in the <a href="thanks-for-help.html">thanks</a> and the modification but it appears this <em>also</em> requires i386 linux; indeed
file) but it appears this <em>also</em> requires i386 linux; indeed looking at the code looking at the code it hard codes paths that are i386 specific to linux.</p>
it hard codes paths that are i386 specific to linux.</p>
<p>Another point of interest is that the author provided unobfuscated versions <p>Another point of interest is that the author provided unobfuscated versions
which might be of value to look at.</p> which might be of value to look at.</p>
<h4 id="aside-why-were-there-changes-made-if-inabiaf">Aside: why were there changes made if INABIAF ?</h4> <h4 id="aside-why-were-there-changes-made-if-inabiaf">Aside: why were there changes made if INABIAF ?</h4>
<p>This is a good question. The reason is we believe it better to fix some obvious <p>This is a good question. The reason is we believe it better to fix some obvious
problems: there were some bugs that would very possibly prevent it from working problems: there were some bugs that would very possibly prevent it from working
even if it was in i386 linux though Yusuke seemed to get it to work in an even if it was in i386 linux though it did seem to work in an emulator, so
emulator so perhaps not. Still it’s better to have the type of the file pointer perhaps not. Still it’s better to have the type of the file pointer correct and
correct and so that the file can at least be opened in modern systems even if so that the file can at least be opened in modern systems even if the compiler
the compiler won’t work there.</p> won’t work there.</p>
<p>Also the supplementary program, which did not work at all, was fixed (by Cody) <p>Also the supplementary program, which did not work at all, was fixed and it can
and it can be run by itself for fun in modern systems, which was not possible be run by itself for fun in modern systems, which was not possible before the
before the fixes there.</p> fixes there.</p>
<div id="2001_cheong"> <div id="2001_cheong">
<h2 id="cheong">2001/cheong</h2> <h2 id="cheong">2001/cheong</h2>
</div> </div>
@@ -2051,9 +2042,7 @@ details and a workaround.</p>
<h3 id="source-code-2001westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/westley/westley.c">2001/westley/westley.c</a></h3> <h3 id="source-code-2001westleywestley.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2001/westley/westley.c">2001/westley/westley.c</a></h3>
<h3 id="information-2001westleyindex.html">Information: <a href="2001/westley/index.html">2001/westley/index.html</a></h3> <h3 id="information-2001westleyindex.html">Information: <a href="2001/westley/index.html">2001/westley/index.html</a></h3>
<p>This function uses <code>gets(3)</code> but it would be ideal if it used <code>fgets(3)</code>. This <p>This function uses <code>gets(3)</code> but it would be ideal if it used <code>fgets(3)</code>. This
one is rather complicated and will be looked at later (some progress was made by one is rather complicated but you are welcome to try and fix this if you wish.</p>
Cody but it is less important than other work) but you are welcome to try
and fix this if you wish.</p>
<h3 id="status-main-has-only-one-arg---try-and-make-it-have-2-or-3">STATUS: main() has only one arg - try and make it have 2 or 3</h3> <h3 id="status-main-has-only-one-arg---try-and-make-it-have-2-or-3">STATUS: main() has only one arg - try and make it have 2 or 3</h3>
<p>This program only has one arg to <code>main()</code>. However changing it to have 2 is not <p>This program only has one arg to <code>main()</code>. However changing it to have 2 is not
as simple as it might seem. Doing this breaks things. If memory serves changing as simple as it might seem. Doing this breaks things. If memory serves changing
@@ -2321,14 +2310,14 @@ for running itself.</p>
<h3 id="status-inabiaf---please-do-not-fix-50">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-50">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-2005mynxmynx.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2005/mynx/mynx.c">2005/mynx/mynx.c</a></h3> <h3 id="source-code-2005mynxmynx.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2005/mynx/mynx.c">2005/mynx/mynx.c</a></h3>
<h3 id="information-2005mynxindex.html">Information: <a href="2005/mynx/index.html">2005/mynx/index.html</a></h3> <h3 id="information-2005mynxindex.html">Information: <a href="2005/mynx/index.html">2005/mynx/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> notes that, though <p>Probably this is obvious, but this entry will not work with https. An
probably obvious, this entry will not work with https. He added an alt version alt version exists that <em>SCANS</em> for https in case someone wants to set up a pipeline or
that scans for https in case someone wants to set up a pipeline or other other workaround (perhaps with <code>stunnel</code>) or even add additional code to it to
workaround (perhaps with stunnel) or even add additional code to it to make it make it work with https but the problem otherwise is a secure connection has to
work with https but the problem otherwise is a secure connection has to first be first be set up in order to give http commands. This is not a bug but it’s worth
set up in order to give http commands. This is not a bug but it’s worth pointing pointing out as it won’t work on as many websites as it used to including the
out as it won’t work on as many websites as it used to including the <a href="https://www.ioccc.org">IOCCC <a href="https://www.ioccc.org">IOCCC website</a> itself. The author noted that someone did
website</a> itself.</p> make a version that supports <code>https</code> but it is not known where this might be.</p>
<div id="2005_sykes"> <div id="2005_sykes">
<h2 id="sykes">2005/sykes</h2> <h2 id="sykes">2005/sykes</h2>
</div> </div>
@@ -2434,13 +2423,12 @@ without outputting a solution. If you can wait that long.</p>
<h3 id="status-inabiaf---please-do-not-fix-57">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3> <h3 id="status-inabiaf---please-do-not-fix-57">STATUS: INABIAF - please <strong>DO NOT</strong> fix</h3>
<h3 id="source-code-2006toledo2toledo2.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2006/toledo2/toledo2.c">2006/toledo2/toledo2.c</a></h3> <h3 id="source-code-2006toledo2toledo2.c">Source code: <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2006/toledo2/toledo2.c">2006/toledo2/toledo2.c</a></h3>
<h3 id="information-2006toledo2index.html">Information: <a href="2006/toledo2/index.html">2006/toledo2/index.html</a></h3> <h3 id="information-2006toledo2index.html">Information: <a href="2006/toledo2/index.html">2006/toledo2/index.html</a></h3>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> fixed this program to <p>This was fixed to
not crash in macOS (it appeared to work fine in fedora linux on an <code>x86_64</code> not crash in macOS (it appeared to work fine in fedora linux on an <code>x86_64</code>
architecture but he does not know if it crashed in macOS Ventura due to the fact architecture but it is not known if it crashed in macOS Ventura due to the fact
it is an <code>arm64</code> CPU - the Apple silicon chip, but given that the author later it is an <code>arm64</code> CPU - the Apple silicon chip, but given that the author later
made some slight changes to make it not crash in <code>x86_64</code> macOS that might be made some slight changes to make it not crash in <code>x86_64</code> macOS that might be
the case). This fix, BTW, is a mix of a fix by <a href="authors.html#Yusuke_Endoh">Yusuke the case).</p>
Endoh</a> and Cody’s.</p>
<p>Nevertheless there are two features that are not bugs.</p> <p>Nevertheless there are two features that are not bugs.</p>
<p>By design this program is supposed to crash on termination. This actually no <p>By design this program is supposed to crash on termination. This actually no
longer seems to happen but we don’t need this to be added even if it might longer seems to happen but we don’t need this to be added even if it might
@@ -2544,11 +2532,10 @@ a bus error).</p>
like it <em>might</em> (see Apple resources below) be but it could also be a wider like it <em>might</em> (see Apple resources below) be but it could also be a wider
problem with macOS. It appears that Apple does not allow certain combinations of problem with macOS. It appears that Apple does not allow certain combinations of
memory protections with the arm64 processor.</p> memory protections with the arm64 processor.</p>
<p><a href="authors.html#Cody_Boone_Ferguson">Cody Boone Ferguson</a> tried some of the <p>Some of the workarounds have been tried but at this time it did not seem to
workarounds but at this time he could not get it to work. He offers some earlier work. Some earlier debugging sessions are below as well as a resource from the
debugging sessions below as well as a resource from the author as well as some author as well as some resources on Apple’s website should anyone wish to take a
resources on Apple’s website should anyone wish to take a crack at it. He might crack at it. A starting point might be in
address it later but a starting point might be in
<a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2011/richards/richards.alt.c">richards.alt.c</a>.</p> <a href="https://github.com/ioccc-src/temp-test-ioccc/blob/master/2011/richards/richards.alt.c">richards.alt.c</a>.</p>
<h3 id="debugging">Debugging</h3> <h3 id="debugging">Debugging</h3>
<p>At first glance it appeared to be that it might be the function pointers or the <p>At first glance it appeared to be that it might be the function pointers or the
@@ -2689,7 +2676,7 @@ Silicon</a></p>
Entitlement</a></p> Entitlement</a></p>
<p>Looking at <p>Looking at
<a href="https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c" class="uri">https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c</a> <a href="https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c" class="uri">https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c</a>
Cody noticed a curious thing in the <code>mprotect()</code> function: <code>#if CONFIG_DYNAMIC_CODE_SIGNING</code> and the protection variable has the A curious thing is in the <code>mprotect()</code> function: <code>#if CONFIG_DYNAMIC_CODE_SIGNING</code> and the protection variable has the
<code>VM_PROT_TRUSTED</code> set it might allow overriding the problem but this is <code>VM_PROT_TRUSTED</code> set it might allow overriding the problem but this is
unconfirmed and it’s not known when <code>CONFIG_DYNAMIC_CODE_SIGNING</code> would be unconfirmed and it’s not known when <code>CONFIG_DYNAMIC_CODE_SIGNING</code> would be
defined.</p> defined.</p>
@@ -2945,7 +2932,7 @@ is not. Again see his index.html for details. However it is also possible to not
rely on the entry as the below shows. One should be able to do:</p> rely on the entry as the below shows. One should be able to do:</p>
<pre><code> echo &#39;No. I want chocolate!&#39; | ./prog &gt; chocolate.raw <pre><code> echo &#39;No. I want chocolate!&#39; | ./prog &gt; chocolate.raw
./prog e &lt; chocolate.raw</code></pre> ./prog e &lt; chocolate.raw</code></pre>
<p>and get:</p> <p>and get:</p>
<pre><code> No. I want some chocolate!</code></pre> <pre><code> No. I want some chocolate!</code></pre>
<p>but instead we get:</p> <p>but instead we get:</p>
bugs.md
+160 -177
View File
@@ -69,15 +69,14 @@ might seem to be fine but is actually not! We will note some where this is known
to happen (if it's not yet fixed). This has actually already happened to us. to happen (if it's not yet fixed). This has actually already happened to us.
Sometimes it is okay: some entries are known to not work but they can still have Sometimes it is okay: some entries are known to not work but they can still have
improvements (rarer situation but it's happened). Some might have slightly improvements (rarer situation but it's happened). Some might have slightly
different output but which is not a problem for instance a newline after output different output but which is not a problem for instance a newline after output.
in [Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson)'s changing it to This is not the most ideal but it is okay.
[1994/ldb](%%REPO_URL%%/1994/ldb/ldb.c) use `fgets()` from `gets()`.
Make **ABSOLUTE CERTAIN** that you read the index.html file _BEFORE_ your changes Make **ABSOLUTE CERTAIN** that you read the index.html file _BEFORE_ your changes
as it's important to see that the code is doing what it is supposed to. In the as it's important to see that the code is doing what it is supposed to. In the
case that it's not obvious (some index.html files do not even have commands to case that it's not obvious (some index.html files do not even have commands to
try or even explain how to use it!) then you should either skip the entry OR ask try or even explain how to use it!) then you should either skip the entry OR ask
someone who will know e.g. us. someone who will know for instance the Judges.
Make **ABSOLUTE CERTAIN** that you read the next section, the list of statuses Make **ABSOLUTE CERTAIN** that you read the next section, the list of statuses
and the related information, BEFORE you submit a pull request! and the related information, BEFORE you submit a pull request!
@@ -124,14 +123,14 @@ Another type of warning that would be hard to keep track of is different data
sizes on different platforms. These tend to be required at the risk that sizes on different platforms. These tend to be required at the risk that
sometimes the entry will not work for certain platforms, some of which might or sometimes the entry will not work for certain platforms, some of which might or
might not be fixable; a good example where it was required to change and is okay might not be fixable; a good example where it was required to change and is okay
is when [Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed the is the fix of the segfault in macOS of
segfault in macOS of [1989/paul](%%REPO_URL%%/1989/paul/paul.c): changing the `int *` to a [1989/paul](%%REPO_URL%%/1989/paul/paul.c): changing the `int *` to a `long *`
`long *` was required and it works just as well with linux. was required and it works just as well with linux.
But even if they are fixable (which will likely be hard to do) it's almost But even if they are fixable (which will likely be hard to do) it's almost
certain that such code would be just as non-portable (importable ? :-) ). certain that such code would be just as non-portable (importable ? :-) ).
In rare cases they are valid. For instance Cody noticed in linux the following In rare cases they are valid. For instance in linux the following
warning in [1985/applin](%%REPO_URL%%/1985/applin/applin.c) with gcc: warning in [1985/applin](%%REPO_URL%%/1985/applin/applin.c) with gcc:
``` ```
@@ -140,7 +139,7 @@ warning in [1985/applin](%%REPO_URL%%/1985/applin/applin.c) with gcc:
| ^~~~~~ | ^~~~~~
``` ```
and this reminded him of his fix to and this is is kind of like the fix to
[1984/anonymous](%%REPO_URL%%/1984/anonymous/anonymous.c). In the case of `1984/anonymous` it [1984/anonymous](%%REPO_URL%%/1984/anonymous/anonymous.c). In the case of `1984/anonymous` it
was a bit more involved but with `1985/applin` one need only add to the Makefile was a bit more involved but with `1985/applin` one need only add to the Makefile
`-include unistd.h`. So there are some cases where fixing warnings can fix a `-include unistd.h`. So there are some cases where fixing warnings can fix a
@@ -164,7 +163,7 @@ include you can update the Makefile `CINCLUDE` variable. For instance if it
needs `stdio.h` you could do `-include stdio.h`. Please leave a space after the needs `stdio.h` you could do `-include stdio.h`. Please leave a space after the
`=` in the Makefile. You may also have extra long lines if this seems useful to `=` in the Makefile. You may also have extra long lines if this seems useful to
make it a one-liner even if it kind of makes it longer than what the judges make it a one-liner even if it kind of makes it longer than what the judges
consider a one-liner. Thank you! consider a one-liner, at least within reason. Thank you!
### On layout of program source: ### On layout of program source:
@@ -172,9 +171,9 @@ consider a one-liner. Thank you!
If you make changes _PLEASE TRY AND KEEP THE SOURCE CODE LAYOUT AS CLOSE TO THE If you make changes _PLEASE TRY AND KEEP THE SOURCE CODE LAYOUT AS CLOSE TO THE
ORIGINAL AS POSSIBLE_. This might not always be possible and if you have an ORIGINAL AS POSSIBLE_. This might not always be possible and if you have an
editor that does formatting it can cause problems. Sometimes formatters can even editor that does formatting it can cause problems. Sometimes formatters can even
break code! [Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) has break code! Some have experienced this many times with vim so he tends to
experienced this many times with vim so he tends to disable all format options disable all format options when formatting code. In vim you should be able to do
when formatting code. In vim you should be able to do that with: that with:
``` ```
:set formatoptions= :set formatoptions=
@@ -261,7 +260,7 @@ compiles but two versions of code it generates does not work with clang.
However some still might have problems of some kind or another and some might However some still might have problems of some kind or another and some might
not yet be located as an additional defect of clang was noticed where `main()` not yet be located as an additional defect of clang was noticed where `main()`
is not allowed four args or one arg. is not allowed four args or one arg (see below).
This is what the status means, however. This is what the status means, however.
@@ -272,26 +271,26 @@ versions have an additional defect where they only allow 0, 2 or 3 args, the
latter of which affected `1989/westley`, so these entries do not work with latter of which affected `1989/westley`, so these entries do not work with
clang. clang.
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) looked at the source Looking at the source
code of clang and reported that there is no way to override the requirement of code of clang, it appears that there is no way to override the requirement of
arg types so these entries will fail to compile with clang. That's why in some arg types so these entries will fail to compile with clang. That's why in some
entries he fixed he did it by adding a function called `pain()`. :-) He did not entries the fix was done with a new function, often called `pain()` :-)
check for the number of args as that was discovered later but he suspects it's
the same issue.
**NOTE for macOS users**: please be aware that _gcc_ under macOS **is actually **NOTE for macOS users**: please be aware that _gcc_ under macOS **is actually
clang** despite the fact it might appear to be gcc: no symlink and both gcc and clang** despite the fact it might appear to be gcc: no symlink and both gcc and
clang exist but the gcc is clang which you'll see if you run `gcc --version`. clang exist but the gcc is clang which you'll see if you run `gcc --version`.
A tip and some fix methods from Cody: in the older days args to main() not given A tip and some fix methods from [Cody Boone
a type were implicit ints but when they're required to be `char **` this can Ferguson](authors.html#Cody_Boone_Ferguson): in the older days args to main()
cause a problem. In some cases Cody was able to use a `char *` inside `main()` not given a type were implicit ints but when they're required to be `char **`
(see [1989/tromp/tromp.c](%%REPO_URL%%/1989/tromp/tromp.c) and this can cause a problem. In some cases it was possible to use a `char *` inside
[1986/holloway/holloway.c](1986/holloway/index.html) for two examples though done slightly `main()` (see [1989/tromp/tromp.c](%%REPO_URL%%/1989/tromp/tromp.c) and
differently). In other cases he was able to dereference the pointers to be used [1986/holloway/holloway.c](1986/holloway/index.html) for two examples though
like an int and other times a cast was necessary. He used other various done slightly differently). In other cases it was possible to dereference the
techniques to get them to compile. In some cases this introduced a problem but pointers to be used like an int and other times a cast was necessary. There are
typically if not always that problem exists with compilers that are less strict. various techniques to get these to compile. In some cases this introduced a
problem but typically if not always that problem exists with compilers that are
less strict.
## STATUS: main() has only one arg ## STATUS: main() has only one arg
@@ -327,11 +326,9 @@ or if you can `-ggdb3` when compiling as that will help with debugging symbols.
**Please help us change use of gets() to fgets()**, if possible. **Please help us change use of gets() to fgets()**, if possible.
Entries with this status use `gets()` which is unsafe because it has no limit on Entries with this status use `gets()` which is unsafe because it has no limit on
the length of the string. [Cody Boone the length of the string. Many were fixed but not all.
Ferguson](authors.html#Cody_Boone_Ferguson) has fixed several of these and he's
working on others in between fixing other things.
There's a very good reason for this even though it is bordering on tampering There's a very good reason for this change even though it is bordering on tampering
entries: in modern systems one will get possibly alarming and certainly annoying entries: in modern systems one will get possibly alarming and certainly annoying
warnings during at least one of compiling, linking and execution. In macOS it's warnings during at least one of compiling, linking and execution. In macOS it's
most annoying as it happens during execution (also compilation) with the warning most annoying as it happens during execution (also compilation) with the warning
@@ -339,30 +336,30 @@ being interspersed with the output of the program, often causing confusing
output with the entry. A good example that was fixed is output with the entry. A good example that was fixed is
[1990/tbr](%%REPO_URL%%/1990/tbr/tbr.c). [1990/tbr](%%REPO_URL%%/1990/tbr/tbr.c).
Cody provided some important notes on this status with respect to fixing / Some important notes on this status with respect to fixing / changing the
changing entries: entries:
0. In some cases changing the entry to use `fgets()` will break the entry and in 0. In some cases changing the entry to use `fgets()` will break the entry and in
some cases cause it to crash. Where this is known we will document it. Some of some cases cause it to crash. Where this is known we will document it.
these Cody is also off and on working on.
1. You **MUST** check the output before and after to make sure that it remains 1. You **MUST** check the output before and after to make sure that it remains
the same. Sometimes the output might not be immediately obviously wrong but is. the same. Sometimes the output might not be immediately obviously wrong but is.
We will document known examples as we come across them. We will document known examples as we come across them.
2. `gets()` does NOT STORE the `'\n'` but `fgets()` DOES! This is one of the 2. `gets()` does NOT STORE the `'\n'` but `fgets()` DOES! This is one of the
reasons some of the entries break when changed to `fgets()`. Cody has fixed some reasons some of the entries break when changed to `fgets()`. Some of those have
of these but there are others he hasn't had time to address or even discover been fixed but there are others that have not been.
yet.
3. Just because you don't see the string `gets` in the code does not mean it's 3. Just because you don't see the string `gets` in the code does not mean it's
not used. It is after all the International _Obfuscated C Code_ Contest! :-) not used. It is after all the International _Obfuscated C Code_ Contest! :-)
Sometimes it will be in the Makefile and other times it will be obfuscated in Sometimes it will be in the Makefile and other times it will be obfuscated in
other ways. Compilers and linkers tend to warn about its use (and as noted in other ways; a good example of that that was fixed is
macOS it also happens at execution) and this is a good way to find entries that [1991/dds](1991/dds/index.html): in this case the string was encrypted.
use it even if it's not visible in the code. Compilers and linkers tend to warn about its use (and as noted in macOS it also
happens at execution) and this is a good way to find entries that use it even if
it's not visible in the code.
_NOTE_: this status is _NOT necessarily mutually exclusive_ with the _INABIAF_ **NOTE**: this status is _NOT necessarily mutually exclusive_ with the _INABIAF_
(it's not a bug it's a feature) status. The reason for this is due to warnings (it's not a bug it's a feature) status. The reason for this is due to warnings
during compiling, linking and/or runtime, sometimes causing confusing output (as during compiling, linking and/or runtime, sometimes causing confusing output (as
noted above). noted above).
@@ -421,7 +418,7 @@ system specific winning entries were discouraged. This is an all time personal
favourite of Landon Curt Noll. Run the alternate code to understand why this favourite of Landon Curt Noll. Run the alternate code to understand why this
might be (along with how strange the source code is :-) ). might be (along with how strange the source code is :-) ).
An example where a crash is not a bug: [2019/endoh](%%REPO_URL%%/2019/endoh/prog.c) is An example where a crash is not a bug: [2019/endoh](2019/endoh/index.html) is
supposed to crash. There are others that are also supposed to crash or that are supposed to crash. There are others that are also supposed to crash or that are
known to segfault but are considered features. known to segfault but are considered features.
@@ -437,7 +434,7 @@ felt that the fix was tampering with the entry).
**NOTE**: in the case of `gets()` we've fixed some to avoid the warning of the **NOTE**: in the case of `gets()` we've fixed some to avoid the warning of the
compiler, linker or even during runtime, depending on the system. In [some cases compiler, linker or even during runtime, depending on the system. In [some cases
like 1990/tbr](%%REPO_URL%%/1990/tbr/tbr.c) the fix actually prevents confusing output (though that like 1990/tbr](1990/tbr/index.html) the fix actually prevents confusing output (though that
was not the only fix made in that entry). was not the only fix made in that entry).
@@ -499,17 +496,18 @@ This program will very likely crash or do something funny without an arg.
### Source code: [1984/mullender/mullender.c](%%REPO_URL%%/1984/mullender/mullender.c) ### Source code: [1984/mullender/mullender.c](%%REPO_URL%%/1984/mullender/mullender.c)
### Information: [1984/mullender/index.html](1984/mullender/index.html) ### Information: [1984/mullender/index.html](1984/mullender/index.html)
Although there is an alt version and supplementary program added by Cody that Although there is an alt version and supplementary program added, that
will work in modern systems, if you do not have a will work in modern systems, if you do not have a
[VAX-11](https://en.wikipedia.org/wiki/VAX-11) or [VAX-11](https://en.wikipedia.org/wiki/VAX-11) or
[PDP-11](https://en.wikipedia.org/wiki/PDP-11) to run the original entry on it [PDP-11](https://en.wikipedia.org/wiki/PDP-11) to run the original entry on it
will not work. See the index.html for details on the alternate versions. will not work. See the index.html for details on the alternate versions.
Cody added and fixed the [gentab.c](%%REPO_URL%%/1984/mullender/gentab.c) which is from the The file [gentab.c](%%REPO_URL%%/1984/mullender/gentab.c) which is from the
author's (or one of them, Mullender) remarks found by Cody. Cody fixed this to author's (or one of them, Mullender) remarks, was fixed for modern systems, at
compile and work (as best as he can determine: he has no VAX-11 or PDP-11 or least as best as can be determined: running the code on the binary itself does
emulator to test it) but running the code on the binary itself produces a produce a `short[]` that can compile in modern systems but it will not work.
`short[]` that can compile in modern systems though it'll not work. Before the fix it would fail in some cases and it is possible that in some cases
it might still fail.
<hr style="width:10%;text-align:left;margin-left:0"> <hr style="width:10%;text-align:left;margin-left:0">
@@ -630,9 +628,11 @@ reverse of other code (also wrt names). See the source file and the index.html
Fixing the (mis)feature is likely to be a very difficult challenge especially Fixing the (mis)feature is likely to be a very difficult challenge especially
without breaking something else which is far more likely (see below in tips from without breaking something else which is far more likely (see below in tips from
Cody). You are welcome to try and fix it if you can! Cody, who fixed it so the original would compile with `clang` and at least one
generated version would compile with `clang` too, not compromising the other
versions with `gcc`). You are welcome to try and fix it if you can!
### Tips from Cody: ### Tips:
The reason this is crashing is that the array `irk` is being accessed way out of The reason this is crashing is that the array `irk` is being accessed way out of
bounds by the int `gnat`. For instance: bounds by the int `gnat`. For instance:
@@ -651,12 +651,11 @@ bounds by the int `gnat`. For instance:
The real trouble is that the code is generated and in a complex way or rather The real trouble is that the code is generated and in a complex way or rather
ways. ways.
I will explain this a little bit (though it might be more than a bit according Trying to explain this might be tricky but here is an attempt. The program, when
to many :-) ) but that is all I can do for now. The program, when used without used without any args, will print out what it reads. If one arg is specified it
any args, will print out what it reads. If one arg is specified it will do a will do a ROT13 of it. If two args are specified it will print out the text
ROT13 of it. If two args are specified it will print out the text backwards backwards (_WITHOUT ROT13!_). If three args are specified it will do both
(_WITHOUT ROT13!_). If three args are specified it will do both **_reversal AND **_reversal AND ROT13_**.
ROT13_**.
So `ver0` is the same as the main program; `ver1` is the ROT13 of the code (but So `ver0` is the same as the main program; `ver1` is the ROT13 of the code (but
see below); `ver2` will reverse the code and `ver3` will both reverse and ROT13 see below); `ver2` will reverse the code and `ver3` will both reverse and ROT13
@@ -714,20 +713,18 @@ and then another version that has the reversal _and_ ROT13 then you can see it's
nigh impossible to fix if not impossible. nigh impossible to fix if not impossible.
The fix to get it to even compile with clang and then get the first two The fix to get it to even compile with clang and then get the first two
(counting the original, `ver0`) to work was the hardest fix I made due to how (counting the original, `ver0`) to work was quite hard due to how the comments
the comments in reverse and ROT13 of it all works together, especially allowing in reverse and ROT13 of it all works together, especially allowing all versions
all versions to compile with gcc. The only reason that it doesn't work to compile with gcc. The only reason that it doesn't work completely with clang
completely with clang is the arg types of main() (see if you can figure out how is the arg types of main() (see if you can figure out how the change of `main()`
the change of `main()` in `westley.c` is not carried over to versions 2 and 3!); in `westley.c` is not carried over to versions 2 and 3!); originally none of
originally none of them compiled with clang. them compiled with clang.
But it was possible to get the ROT13 code to compile with clang but the other But it was possible to get the ROT13 code to compile with clang but the other
versions, reversed as well as ROT13 and reversed code, proved much more versions, reversed as well as ROT13 and reversed code, proved much more
problematic. I got it to compile (or maybe mostly compile) all versions with problematic. It is now possible to compile (or maybe mostly compile) all versions with
clang but I then carried it over to gcc in another system and it had syntax clang but on another system and it had syntax errors.
errors.
That's about all I can say for how it works as other things have to be done too.
Enjoy! :-) Enjoy! :-)
@@ -819,11 +816,10 @@ This is not considered a bug, however.
### Source code: [1990/jaw/jaw.c](%%REPO_URL%%/1990/jaw/jaw.c) ### Source code: [1990/jaw/jaw.c](%%REPO_URL%%/1990/jaw/jaw.c)
### Information: [1990/jaw/index.html](1990/jaw/index.html) ### Information: [1990/jaw/index.html](1990/jaw/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) Some issues in this program were fixed and [Yusuke
fixed some issues in this program and [Yusuke Endoh](authors.html#Yusuke_Endoh) then provided the `btoa` script but it appears
Endoh](authors.html#Yusuke_Endoh) provided the `btoa` there is a bug in this entry. The judges wrote that to test the entry one can
script but it appears there is a bug in this entry. The judges wrote that to do:
test the entry one can do:
``` <!---sh--> ``` <!---sh-->
echo "Quartz glyph jocks vend, fix, BMW." | compress | ./btoa | ./jaw echo "Quartz glyph jocks vend, fix, BMW." | compress | ./btoa | ./jaw
@@ -889,11 +885,11 @@ The authors provided a list of features in the
### Source code: [1990/theorem/theorem.c](%%REPO_URL%%/1990/theorem/theorem.c) ### Source code: [1990/theorem/theorem.c](%%REPO_URL%%/1990/theorem/theorem.c)
### Information: [1990/theorem/index.html](1990/theorem/index.html) ### Information: [1990/theorem/index.html](1990/theorem/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed many bugs that Many bugs that prevented this from working properly (including segfaults) were
prevented this from working properly (including segfaults) but one thing to note fixed but one thing to note is that if you pass two zeroes to `theorem_bkp` or
is that if you pass two zeroes to `theorem_bkp` or `fibonacci` the program `fibonacci` the program will enter an infinite loop, printing 0 over and over
will enter an infinite loop, printing 0 over and over again; another condition again; another condition where this occurred was fixed but this one should not
where this occurred was fixed but this one should not be fixed. Thank you. be fixed.
<div id="1990_westley"> <div id="1990_westley">
@@ -904,7 +900,7 @@ where this occurred was fixed but this one should not be fixed. Thank you.
### Source code: [1990/westley/westley.c](%%REPO_URL%%/1990/westley/westley.c) ### Source code: [1990/westley/westley.c](%%REPO_URL%%/1990/westley/westley.c)
### Information: [1990/westley/index.html](1990/westley/index.html) ### Information: [1990/westley/index.html](1990/westley/index.html)
Although Cody fixed this to not enter an infinite loop if the arg (converted to Although this was fixed to not enter an infinite loop if the arg (converted to
a number) is < 0 the lack of an arg check at all was kept in to make it like the a number) is < 0 the lack of an arg check at all was kept in to make it like the
original. The reason for the < 0 check is it floods the screen (yes this is original. The reason for the < 0 check is it floods the screen (yes this is
indeed inconsistent with some other entries but it is not worth the time spent, indeed inconsistent with some other entries but it is not worth the time spent,
@@ -969,9 +965,8 @@ when you're cheating it ends up winning! Can you figure that out as well?
### Information: [1992/adrian/index.html](1992/adrian/index.html) ### Information: [1992/adrian/index.html](1992/adrian/index.html)
The author stated that if the file cannot be opened then it will print a system The author stated that if the file cannot be opened then it will print a system
error but this is not the case unless it's showing a fault. However this was error but this is not the case unless it's showing a fault. This was fixed as
instead fixed by Cody as part of another problem that was detected so that the part of another problem that was detected.
program just exits 1 if the file cannot be opened.
Also, if _some_ of the programs are not given an arg they will very likely crash Also, if _some_ of the programs are not given an arg they will very likely crash
or do something funny. For instance `adsleep` and `adbasename`. This is very or do something funny. For instance `adsleep` and `adbasename`. This is very
@@ -1022,7 +1017,7 @@ Leo Broukhis, before he was an IOCCC judge, sent the IOCCC judges an email:
A quick debugging session from Cody suggests that the problem with this value A quick debugging session suggests that the problem with this value
is that: is that:
``` <!---c--> ``` <!---c-->
@@ -1105,7 +1100,7 @@ debate.
### Information: [1992/kivinen/index.html](1992/kivinen/index.html) ### Information: [1992/kivinen/index.html](1992/kivinen/index.html)
When you start the program everything starts to move over to the right side and When you start the program everything starts to move over to the right side and
then ends. [Yusuke Endoh](authors.html#Yusuke_Endoh) pointed out that if you then ends. [Yusuke Endoh](authors.html#Yusuke_Endoh) pointed out that if you
click the mouse it takes it back towards the centre. click the mouse it takes it back towards the centre.
If you want to try and fix this, you are welcome to try. If you want to try and fix this, you are welcome to try.
@@ -1119,15 +1114,15 @@ If you want to try and fix this, you are welcome to try.
### Source code: [1992/lush/lush.c](%%REPO_URL%%/1992/lush/lush.c) ### Source code: [1992/lush/lush.c](%%REPO_URL%%/1992/lush/lush.c)
### Information: [1992/lush/index.html](1992/lush/index.html) ### Information: [1992/lush/index.html](1992/lush/index.html)
We used a patch from [Yusuke Endoh](authors.html#Yusuke_Endoh) to get this to This was fixed but it only properly works with gcc; it was also changed to use
work but it only works with gcc. Cody removed the warnings of `gets()` and `fgets()` instead of `gets()` and a script was provided to run the entry
provided a script to run the entry properly. properly. Nonetheless it will only work with gcc.
Unfortunately due to the way the entry works and the fact that other compilers Unfortunately due to the way the entry works and the fact that other compilers
like clang have different warnings and errors this simply does not work with like clang have different warnings and errors this simply does not work with
them. them.
Some tips from Cody: Some tips:
This entry relies on specific compiler warnings. With gcc it will look something This entry relies on specific compiler warnings. With gcc it will look something
like: like:
@@ -1217,7 +1212,7 @@ it yourself through ctrl-c or killing it in some other fashion.
### Source code: [1992/westley/westley.c](%%REPO_URL%%/1992/westley/westley.c) ### Source code: [1992/westley/westley.c](%%REPO_URL%%/1992/westley/westley.c)
### Information: [1992/westley/index.html](1992/westley/index.html) ### Information: [1992/westley/index.html](1992/westley/index.html)
Cody improved the usability of this program by making it so that as long as the The usability of this program was improved by making it so that as long as the
terminal columns is >= 80 it will display properly, rather than having to wrap terminal columns is >= 80 it will display properly, rather than having to wrap
at 80 columns. However due to the nature of the program if the terminal is < 80 at 80 columns. However due to the nature of the program if the terminal is < 80
in column width it will not display right. To see the number of columns in your in column width it will not display right. To see the number of columns in your
@@ -1392,13 +1387,13 @@ there.
### Source code: [1994/ldb/ldb.c](%%REPO_URL%%/1994/ldb/ldb.c) ### Source code: [1994/ldb/ldb.c](%%REPO_URL%%/1994/ldb/ldb.c)
### Information: [1994/ldb/index.html](1994/ldb/index.html) ### Information: [1994/ldb/index.html](1994/ldb/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed this to compile This was fixed to compile with modern systems but the entry also used `gets(3)`
with modern systems but the entry also used `gets(3)` which in some systems would which in some systems would print out a warning along with the output of the
print out a warning along with the output of the program. Naturally it could program. Naturally it could also overflow long lines, of which there are many,
also overflow long lines, of which there are many, though that is more though that is more considered a feature, not a bug, as it was documented by the
considered a feature, not a bug, as it was documented by the author. author.
Cody changed it to `fgets(3)` to prevent the display problem but this introduces It was changed to use `fgets(3)` to prevent the display problem but this introduces
another problem namely that newlines can be printed if the line length < 231. another problem namely that newlines can be printed if the line length < 231.
This seems like a worthy compromise to not have messed up output and although it This seems like a worthy compromise to not have messed up output and although it
@@ -1423,7 +1418,7 @@ for more information on the change to `fgets(3)`.
even when it works. See the author's remarks in the index.html for details. even when it works. See the author's remarks in the index.html for details.
Increasing the buffer size and having it, when fed its own source code, Increasing the buffer size and having it, when fed its own source code,
generate code that will compile with the same buffer size is difficult: it just generate code that will compile with the same buffer size is difficult: it just
uses the original size. Cody explains this further down. uses the original size. This is explained later down.
Getting it to use `fgets()` is even harder as when fed its own code it will Getting it to use `fgets()` is even harder as when fed its own code it will
generate code that cannot even compile let alone use `fgets(3)`. generate code that cannot even compile let alone use `fgets(3)`.
@@ -1432,37 +1427,31 @@ In both cases, changing the buffer size and changing it to use `fgets(3)`, will
cause compilation errors without other adjustments. cause compilation errors without other adjustments.
See the below magic for details. Along with the author's remarks in the See the below magic for details. Along with the author's remarks in the
index.html file it might prove possible to get it to use `fgets(3)`. Cody, index.html file it might prove possible to get it to use `fgets(3)`.
writing this on 02 November 2023, only just noticed the author's remarks and
will later on look at this if nobody takes up the challenge. More important work
like getting to a place that the next contest can run must be done first.
Cody explains the magic of how this entry works, which will be necessary if this Here is the magic of how this entry works, to help anyone who wishes to work on
entry is to be fixed, below. this.
You might wish to run the command:
``` <!---sh-->
make diff_orig_prog
```
to see what had to change for the buffer size, when looking at the below.
Furthermore you will want to look at:
You might wish to run the commands:
``` <!---sh--> ``` <!---sh-->
diff schnitzi.alt.c schnitzi.alt2.c diff schnitzi.alt.c schnitzi.alt2.c
diff schnitzi.alt2.c schnitzi.alt.c
``` ```
as the `fgets(3)` version can compile it just can't generate compilable code (let to see how the buffer size and how the `gets(3)` was changed to use `fgets()`,
alone using `fgets(3)`) when fed itself (to reiterate, just changing the call to in the two alternate versions. Notice that the `fgets(3)` version can compile it
`fgets(3)` does not mean it can compile and there's a further problem in that just can't **generate** compilable code (let alone using `fgets(3)`) when fed
`fgets(3)` retains the newline whereas `gets(3)` does not). Nevertheless looking itself (to reiterate, just changing the call to `fgets(3)` does not mean it can
at these commands will be of help to understand how it works. compile and there's a further problem in that `fgets(3)` retains the newline
whereas `gets(3)` does not). Nevertheless looking at these commands will be of
help to understand how it works, in case one wishes to try and fix it.
For the alternate versions the other functionality is unaffected. For the alternate versions the other functionality is unaffected.
### The magic of [1994/schnitzi](%%REPO_URL%%/1994/schnitzi/schnitzi.c) and how it flips text
### The magic of [1994/schnitzi](1994/schnitzi/index.html) and how it flips text
The problem is getting the generated code to use `fgets()` (once it even The problem is getting the generated code to use `fgets()` (once it even
compiles which was easy to do) and also have the updated buffer size be the compiles which was easy to do) and also have the updated buffer size be the
@@ -1524,9 +1513,9 @@ changing the buffer size, see below).
Getting this entry to use `fgets(3)` is easy but the problem is you're supposed Getting this entry to use `fgets(3)` is easy but the problem is you're supposed
to be able to feed the source to the program and the output of that will be to be able to feed the source to the program and the output of that will be
compilable. It however will create compiler errors. So it's not just changing compilable. It however will create compiler errors. So it's not just changing
the code to get it to use `fgets(3)`! I (Cody) already did this before I noticed the code to get it to use `fgets(3)`! This was done but
the other part which is why I rolled that part back. I did increase the buffer the other part is why that part was rolled back. The buffer size was increased
size but that only works on the original source. but that only works on the original source.
To say just how sensitive this entry is: even a space, lack of space or a To say just how sensitive this entry is: even a space, lack of space or a
character, wrong character, lack of character or character in the wrong place character, wrong character, lack of character or character in the wrong place
@@ -1553,7 +1542,7 @@ that's compiled) cannot be opened in the directory it is run from.
### Important reminder and a note about the `-1` value check for `getc()`: ### Important reminder and a note about the `-1` value check for `getc()`:
Cody fixed the code to not use `-1` for the return value of `getc()`; this is This code was fixed to not use `-1` for the return value of `getc()`; this is
important because `EOF` is **NOT** guaranteed to be `-1` but rather any negative important because `EOF` is **NOT** guaranteed to be `-1` but rather any negative
value. On systems where `EOF != -1` the program would enter an infinite loop value. On systems where `EOF != -1` the program would enter an infinite loop
until the program crashed, by chance reads a `-1` or was killed (it is for this until the program crashed, by chance reads a `-1` or was killed (it is for this
@@ -1619,8 +1608,7 @@ as documented, are considered features.
### Information: [1995/cdua/index.html](1995/cdua/index.html) ### Information: [1995/cdua/index.html](1995/cdua/index.html)
This did not originally compile under macOS and after it did compile under This did not originally compile under macOS and after it did compile under
macOS, it crashed. [Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) macOS, it crashed. These problems were fixed.
fixed these problems.
It should be noted however that there is a condition where the program will It should be noted however that there is a condition where the program will
prompt you to press return again. This was thought to be a bug but looking at prompt you to press return again. This was thought to be a bug but looking at
@@ -1988,7 +1976,7 @@ Another feature that does work in linux is:
Can you help fix these problems? Can you help fix these problems?
Cody provided the C code that creates an unformatted perl script that will just Here is some C code that creates an unformatted perl script that will just
print 'unimplemented' for operators but it seems to work in some other ways (see print 'unimplemented' for operators but it seems to work in some other ways (see
below for how to go further than unimplemented) which is the exact perl from the below for how to go further than unimplemented) which is the exact perl from the
C program except that it has the `#!/usr/bin/env perl` and `use warnings;` as C program except that it has the `#!/usr/bin/env perl` and `use warnings;` as
@@ -2097,7 +2085,7 @@ That means then that in linux despite the subroutines seemingly not existing it
works in the code, at least in some cases. works in the code, at least in some cases.
As far as the unimplemented error goes: the author stated that it means there is As far as the unimplemented error goes: the author stated that it means there is
invalid input. But as can be seen by the input I (Cody) gave the input is not invalid input. But as can be seen by the input given, the input is not
actually incorrect. It is also curious to note that it ends up printing what's actually incorrect. It is also curious to note that it ends up printing what's
on the stack at that point: on the stack at that point:
@@ -2107,15 +2095,14 @@ on the stack at that point:
5 5
``` ```
which might (?) suggest that the `+` operator is unimplemented. Unfortunately it which might (?) suggest that the `+` operator is unimplemented.
has been many years since I have used perl and I was never a guru either.
### STATUS: INABIAF - please **DO NOT** fix ### STATUS: INABIAF - please **DO NOT** fix
The author states that in perl < 5.6.0 there is a bug with a core dump in what The author states that in perl < 5.6.0 there is a bug with a core dump in what
they said is in `Perl_sv_upgrade`. As this is documented it is not considered a they said is in `Perl_sv_upgrade`. As this is documented it is not considered a
bug to be fixed. For the curious this will crash in macOS. Cody notes that the bug to be fixed. For the curious this will crash in macOS. It was noted that the
code that crashes is: code that crashes is:
``` <!---c--> ``` <!---c-->
@@ -2224,17 +2211,18 @@ fire](https://en.wikipedia.org/wiki/Halt_and_Catch_Fire_&lpar;computing&rpar;)!
### Source code: [2001/anonymous/anonymous.c](%%REPO_URL%%/2001/anonymous/anonymous.c) ### Source code: [2001/anonymous/anonymous.c](%%REPO_URL%%/2001/anonymous/anonymous.c)
### Information: [2001/anonymous/index.html](2001/anonymous/index.html) ### Information: [2001/anonymous/index.html](2001/anonymous/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed this so that it This was fixed so that it doesn't segfault and then also fixed the functionality
doesn't segfault and then also fixed the functionality of it (but see below). of it (but see below).
Cody also fixed the [anonymous.ten.c](%%REPO_URL%%/2001/anonymous/anonymous.ten.c) program The [anonymous.ten.c](%%REPO_URL%%/2001/anonymous/anonymous.ten.c) program was
(it also segfaulted) but notes that it **MUST** be compiled as a **32-bit ELF also fixed (it also segfaulted) but note that it **MUST** be compiled as a **32-bit ELF
binary** so it will NOT work in macOS (for example); if you cannot use `-m32` or binary** so it will NOT work in macOS (for example); if you cannot use `-m32` or
even if you can but it cannot compile as an ELF binary (not elf binary :-) ) even if you can but it cannot compile as an ELF binary (not elf binary :-) )
(but see below) then the `anonymous` program itself will very likely crash if (but see below) then the `anonymous` program itself will very likely crash if
you run it on the compiled [anonymous.ten.c](%%REPO_URL%%/2001/anonymous/anonymous.ten.c) but you run it on the compiled
if nothing else it will not modify the target executable (this part of the fix [anonymous.ten.c](%%REPO_URL%%/2001/anonymous/anonymous.ten.c); if nothing else
at least should be correct). it will not modify the target executable (this part of the fix at least should
be correct).
Note also that if you don't specify a file or you specify a non-32-bit ELF file Note also that if you don't specify a file or you specify a non-32-bit ELF file
this program will very likely crash or do something strange like slaughter the this program will very likely crash or do something strange like slaughter the
@@ -2252,28 +2240,27 @@ elves of Imladris :-(
The author stated that it only works with i386 linux so the fact it doesn't work The author stated that it only works with i386 linux so the fact it doesn't work
in modern systems is considered a feature and not a bug. in modern systems is considered a feature and not a bug.
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed an initial An initial segfault was fixed (trying to open the file) and the [supplementary
segfault (trying to open the file) and he also fixed the [supplementary program program bellard.otccex.c](%%REPO_URL%%/2001/bellard/bellard.otccex.c) was also
bellard.otccex.c](%%REPO_URL%%/2001/bellard/bellard.otccex.c) but this still crashes in fixed but this still crashes in systems not i386 linux which is what the author
systems not i386 linux which is what the author stated. stated.
The author said that they compiled it with [gcc version The author said that they compiled it with [gcc version
2.95.2](https://ftp.gnu.org/gnu/gcc/gcc-2.95.2/gcc-everything-2.95.2.tar.gz). We 2.95.2](https://ftp.gnu.org/gnu/gcc/gcc-2.95.2/gcc-everything-2.95.2.tar.gz). We
don't know if a certain gcc version is necessary but it might be helpful to don't know if a certain gcc version is necessary but it might be helpful to
download and compile that version to test it - or it might not. download and compile that version to test it - or it might not.
I (Cody) have no i386 system to test this but perhaps this is why I can't get it Yusuke was able to get this to work with `-m32` but only in an emulator. As
to work. Yusuke was able to get this to work with `-m32` but only in an such, it might be that the changes made should be rolled back but at least now
emulator. the supplementary program will work in modern systems.
On the author's [web page for this program](https://bellard.org/otcc/) it is On the author's [web page for this program](https://bellard.org/otcc/) it is
explicitly stated that it requires i386 linux. explicitly stated that it requires i386 linux.
There I found what should be a more portable version which is included as There it was found what should be a more portable version which is included as
[otccelf.c](%%REPO_URL%%/2001/bellard/otccelf.c) (after adding some `#include`s and the [otccelf.c](%%REPO_URL%%/2001/bellard/otccelf.c) (after adding some `#include`s
modification by Yusuke noted in the [thanks](thanks-for-help.html) and the modification but it appears this *also* requires i386 linux; indeed
file) but it appears this *also* requires i386 linux; indeed looking at the code looking at the code it hard codes paths that are i386 specific to linux.
it hard codes paths that are i386 specific to linux.
Another point of interest is that the author provided unobfuscated versions Another point of interest is that the author provided unobfuscated versions
which might be of value to look at. which might be of value to look at.
@@ -2283,14 +2270,14 @@ which might be of value to look at.
This is a good question. The reason is we believe it better to fix some obvious This is a good question. The reason is we believe it better to fix some obvious
problems: there were some bugs that would very possibly prevent it from working problems: there were some bugs that would very possibly prevent it from working
even if it was in i386 linux though Yusuke seemed to get it to work in an even if it was in i386 linux though it did seem to work in an emulator, so
emulator so perhaps not. Still it's better to have the type of the file pointer perhaps not. Still it's better to have the type of the file pointer correct and
correct and so that the file can at least be opened in modern systems even if so that the file can at least be opened in modern systems even if the compiler
the compiler won't work there. won't work there.
Also the supplementary program, which did not work at all, was fixed (by Cody) Also the supplementary program, which did not work at all, was fixed and it can
and it can be run by itself for fun in modern systems, which was not possible be run by itself for fun in modern systems, which was not possible before the
before the fixes there. fixes there.
<div id="2001_cheong"> <div id="2001_cheong">
@@ -2421,9 +2408,7 @@ See also the author's list of bugs [here](2001/schweikh/index.html#bugs).
### Information: [2001/westley/index.html](2001/westley/index.html) ### Information: [2001/westley/index.html](2001/westley/index.html)
This function uses `gets(3)` but it would be ideal if it used `fgets(3)`. This This function uses `gets(3)` but it would be ideal if it used `fgets(3)`. This
one is rather complicated and will be looked at later (some progress was made by one is rather complicated but you are welcome to try and fix this if you wish.
Cody but it is less important than other work) but you are welcome to try
and fix this if you wish.
### STATUS: main() has only one arg - try and make it have 2 or 3 ### STATUS: main() has only one arg - try and make it have 2 or 3
@@ -2804,14 +2789,14 @@ for running itself.
### Source code: [2005/mynx/mynx.c](%%REPO_URL%%/2005/mynx/mynx.c) ### Source code: [2005/mynx/mynx.c](%%REPO_URL%%/2005/mynx/mynx.c)
### Information: [2005/mynx/index.html](2005/mynx/index.html) ### Information: [2005/mynx/index.html](2005/mynx/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) notes that, though Probably this is obvious, but this entry will not work with https. An
probably obvious, this entry will not work with https. He added an alt version alt version exists that _SCANS_ for https in case someone wants to set up a pipeline or
that scans for https in case someone wants to set up a pipeline or other other workaround (perhaps with `stunnel`) or even add additional code to it to
workaround (perhaps with stunnel) or even add additional code to it to make it make it work with https but the problem otherwise is a secure connection has to
work with https but the problem otherwise is a secure connection has to first be first be set up in order to give http commands. This is not a bug but it's worth
set up in order to give http commands. This is not a bug but it's worth pointing pointing out as it won't work on as many websites as it used to including the
out as it won't work on as many websites as it used to including the [IOCCC [IOCCC website](https://www.ioccc.org) itself. The author noted that someone did
website](https://www.ioccc.org) itself. make a version that supports `https` but it is not known where this might be.
<div id="2005_sykes"> <div id="2005_sykes">
@@ -2968,13 +2953,12 @@ without outputting a solution. If you can wait that long.
### Source code: [2006/toledo2/toledo2.c](%%REPO_URL%%/2006/toledo2/toledo2.c) ### Source code: [2006/toledo2/toledo2.c](%%REPO_URL%%/2006/toledo2/toledo2.c)
### Information: [2006/toledo2/index.html](2006/toledo2/index.html) ### Information: [2006/toledo2/index.html](2006/toledo2/index.html)
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) fixed this program to This was fixed to
not crash in macOS (it appeared to work fine in fedora linux on an `x86_64` not crash in macOS (it appeared to work fine in fedora linux on an `x86_64`
architecture but he does not know if it crashed in macOS Ventura due to the fact architecture but it is not known if it crashed in macOS Ventura due to the fact
it is an `arm64` CPU - the Apple silicon chip, but given that the author later it is an `arm64` CPU - the Apple silicon chip, but given that the author later
made some slight changes to make it not crash in `x86_64` macOS that might be made some slight changes to make it not crash in `x86_64` macOS that might be
the case). This fix, BTW, is a mix of a fix by [Yusuke the case).
Endoh](authors.html#Yusuke_Endoh) and Cody's.
Nevertheless there are two features that are not bugs. Nevertheless there are two features that are not bugs.
@@ -3130,11 +3114,10 @@ like it _might_ (see Apple resources below) be but it could also be a wider
problem with macOS. It appears that Apple does not allow certain combinations of problem with macOS. It appears that Apple does not allow certain combinations of
memory protections with the arm64 processor. memory protections with the arm64 processor.
[Cody Boone Ferguson](authors.html#Cody_Boone_Ferguson) tried some of the Some of the workarounds have been tried but at this time it did not seem to
workarounds but at this time he could not get it to work. He offers some earlier work. Some earlier debugging sessions are below as well as a resource from the
debugging sessions below as well as a resource from the author as well as some author as well as some resources on Apple's website should anyone wish to take a
resources on Apple's website should anyone wish to take a crack at it. He might crack at it. A starting point might be in
address it later but a starting point might be in
[richards.alt.c](%%REPO_URL%%/2011/richards/richards.alt.c). [richards.alt.c](%%REPO_URL%%/2011/richards/richards.alt.c).
@@ -3326,7 +3309,7 @@ Entitlement](https://developer.apple.com/documentation/bundleresources/entitleme
Looking at Looking at
<https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c> <https://github.com/apple/darwin-xnu/blob/5394bb038891708cd4ba748da79b90a33b19f82e/bsd/kern/kern_mman.c>
Cody noticed a curious thing in the `mprotect()` function: `#if A curious thing is in the `mprotect()` function: `#if
CONFIG_DYNAMIC_CODE_SIGNING` and the protection variable has the CONFIG_DYNAMIC_CODE_SIGNING` and the protection variable has the
`VM_PROT_TRUSTED` set it might allow overriding the problem but this is `VM_PROT_TRUSTED` set it might allow overriding the problem but this is
unconfirmed and it's not known when `CONFIG_DYNAMIC_CODE_SIGNING` would be unconfirmed and it's not known when `CONFIG_DYNAMIC_CODE_SIGNING` would be
@@ -3667,7 +3650,7 @@ rely on the entry as the below shows. One should be able to do:
``` <!---sh--> ``` <!---sh-->
echo 'No. I want chocolate!' | ./prog > chocolate.raw echo 'No. I want chocolate!' | ./prog > chocolate.raw
./prog e < chocolate.raw ./prog e < chocolate.raw
``` ```
and get: and get: