Merge pull request #1045 from xexyl/2001-bellard-fixes

2001/bellard/Makefile

@@ -132,9 +132,9 @@ all: data ${TARGET}
 	supernova deep_magic magic charon pluto
 
 ${PROG}: ${PROG}.c
-	@echo "WARNING: ${PROG}.c must be compiled as 32-bit to use with the entry itself;"
-	@echo "if you cannot use -m32 this will NOT work!"
 	${CC} ${CFLAGS} -m32 $< -o $@ ${LDFLAGS} || :
+	@echo "WARNING: ${PROG}.c must be compiled as 32-bit to use with the entry itself:"
+	@echo "if you cannot use -m32 this will NOT work! This entry also requires i386 linux."
 
 ${PROG}.otccex: ${PROG}.otccex.c
 	${CC} ${CFLAGS} $< -o $@ ${LDFLAGS}

2001/bellard/README.md

@@ -1,9 +1,6 @@
 # Best abuse of the rules
 
 Fabrice Bellard  
-451 chemin du mas de Matour  
-34790 Grabels  
-France  
 <https://bellard.org>
 
 

2001/bellard/bellard.c

@@ -14,13 +14,15 @@
 #define f ()
 #define J return
 #define l ae(
+#define n e)
 #define u d!=
 #define F int 
 #define y (j)
 #define r m=
 #define t +4
-F d,z,C,h,P,K,ac,q,G,v,Q,R,D,L,W,M;
+F d,z,C,h,P,K,ac,q,G,v,R,D,L,W,M;
-E(int e){
+FILE *Q;
+E(n{
 x D++=e;
 }
 o f{
@@ -134,43 +136,43 @@ x q++=g;
 g=g>>8;
 }
 }
-A(int e){
+A(n{
 F g;
-p e){
+p n{
 g=k e;
 k e=q-e-4;
 e=g;
 }
 }
-s(int g,int e){
+s(g,n{
 l g);
 k q=e;
 e=q;
 q=q t;
 J e;
 }
-H(e){
+H(n{
-s(184,e);
+s(184,n;
 }
-B(e){
+B(n{
-J s(233,e);
+J s(233,n;
 }
-S(int j,int e){
+S(j,n{
 l 1032325);
-J s(132+j,e);
+J s(132+j,n;
 }
-Z(int e){
+Z(n{
 l 49465);
 H(0);
 l 15);
 l e+144);
 l 192);
 }
-N(int j,int e){
+N(j,n{
 l j+131);
-s((e<512)<<7|5,e);
+s((e<512)<<7|5,n;
 }
-T (int j){
+T y{
 F g,e,m,aa;
 g=1;
 a d b 34){
@@ -224,7 +226,7 @@ w f;
 l 89);
 l 392+(e b 256));
 }
-i a e){
+i a n{
 a e b 256)l 139);
 i l 48655);
 q++;
@@ -267,7 +269,7 @@ k r j;
 c;
 a!g){
 e=e t;
-k e=s(232,k e);
+k e=s(232,k n;
 }
 i a g b 1){
 s(2397439,j);
@@ -279,11 +281,11 @@ s(232,g-q-5);
 a j)s(50305,j);
 }
 }
-O (int j){
+O y{
 F e,g,m;
 a j--b 1)T(1);
 i{
-O (j);
+O y;
 r 0;
 p j b C){
 g=d;
@@ -291,17 +293,17 @@ e=z;
 c;
 a j>8){
 r S(e,m);
-O (j);
+O y;
 }
 i{
 l 80);
-O (j);
+O y;
 l 89);
 a j b 4|j b 5){
-Z(e);
+Z(n;
 }
 i{
-l e);
+l n;
 a g b 37)l 146);
 }
 }
@@ -311,7 +313,7 @@ r S(e,m);
 H(e^1);
 B(5);
 A(m);
-H(e);
+H(n;
 }
 }
 }
@@ -322,19 +324,19 @@ U f{
 w f;
 J S(0,0);
 }
-I (int j){
+I y{
 F m,g,e;
 a d b 288){
 c;
 c;
 r U f;
 c;
-I (j);
+I y;
 a d b 312){
 c;
 g=B(0);
 A(m);
-I (j);
+I y;
 A(g);
 }
 i{
@@ -360,7 +362,7 @@ a u 41){
 e=B(0);
 w f;
 B(g-q-5);
-A(e);
+A(n;
 g=e t;
 }
 }
@@ -372,7 +374,7 @@ A(m);
 i a d b 123){
 c;
 ab(1);
-p u 125)I (j);
+p u 125)I y;
 c;
 }
 i{
@@ -389,7 +391,7 @@ i a u 59)w f;
 c;
 }
 }
-ab (int j){
+ab y{
 F m;
 p d b 256|u-1&!j){
 a d b 256){
@@ -431,7 +433,7 @@ k r G;
 }
 }
 }
-int main(int g,char **e){
+int main(int g, char **e){
 Q=stdin;
 a g-->1){
 *e=e[1];
@@ -444,6 +446,6 @@ P V;
 o f;
 c;
 ab(0);
-J(*(int(*)f)k(P+592))(g,e);
+J(*(int(*)f)k(P+592))(g,n;
 }
 

bugs.md

@@ -1139,14 +1139,24 @@ this program will very likely crash or do something strange like slaughter the
 elves of Imladris :-(
 
 ## [2001/bellard](2001/bellard/bellard.c) ([README.md](2001/bellard/README.md))
+## STATUS: INABIAF - please **DO NOT** fix
 ## STATUS: doesn't work with some platforms - please help us fix
 
+The two statuses might seem contradictory but that is a complicated question.
+The author stated that it only works with i386 linux so on the one hand the fact
+it doesn't work in modern systems is considered a feature and not a bug. But on
+the other hand it would be nice if there was an alternate version which worked
+for modern systems. This does seem quite unlikely but some fixes, described
+next, were made.
+
 [Cody Boone Ferguson](/winners.html#Cody_Boone_Ferguson) fixed an initial
-segfault and he also fixed the [supplementary program
+segfault (trying to open the file) and he also fixed the [supplementary program
-bellard.otccex.c](2001/bellard/bellard.otccex.c) but this still crashes. Below
+bellard.otccex.c](2001/bellard/bellard.otccex.c) but this still crashes in
-are some notes about getting this entry to work but the gist of it is that it
+systems not i386 linux which is what the author stated.
-requires i386 linux. Can you fix it for 64-bit linux? We welcome your help! Here
+
-are some notes of interest:
+On the other hand if you do have a fix for 64-bit systems you're welcome to
+provide it as an alternate version. If by chance you have a fix so that it works
+for both 32-bit and 64-bit systems that is also okay.
 
 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
@@ -1154,27 +1164,39 @@ 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.
 
 I (Cody) have no i386 system to test this but perhaps this is why I can't get it
-to work.  Yusuke was able to get this to work with `-m32` but it seems with an
+to work.  Yusuke was able to get this to work with `-m32` but only in an
 emulator.
 
-On the author's [web page for this program](https://bellard.org/otcc/) where it
+On the author's [web page for this program](https://bellard.org/otcc/) it is
-is stated 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
 [otccelf.c](2001/bellard/otccelf.c) (after adding some `#include`s and the
-modification by Yusuke noted in the README.md file) but it appears this also
+modification by Yusuke noted in the README.md file) but it appears this *also*
 requires i386 linux; indeed looking at the code it hard codes paths that are
 i386 specific to linux.
 
 Another point of interest is that the author provided de-obfuscated versions
 which might be of value to look at. I might do that as well but this entry is
-very likely never going to work for 64-bit linux.
+very likely never going to work for 64-bit linux so that's not that likely since
+there are other things that are more important.
 
-Or maybe you have a fix for 64-bit CPUs? You might like to look at the otccelf
+If you have a fix for 64-bit systems this is welcome as an alternate version, as
-version but note that it at least in 64-bit linux (and macOS) have compilation
+stated above. You might like to look at the otccelf version but note that it (at
-errors.
+least in 64-bit linux and macOS) has compilation errors.
 
-Either way we welcome your help! Thank you!
+### Aside: why were there changes if INABIAF ?
+
+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
+even if it was in i386 linux though Yusuke seemed to get it to work in an
+emulator so perhaps not. Still it's better to have the type of the file pointer
+correct and so that the file can at least be opened in modern systems even if
+the compiler won't work there.
+
+Also the supplementary program, which did not work at all, was fixed (by Cody)
+and it can be run by itself for fun in modern systems, which was not possible
+before the fixes there.
 
 
 ## [2001/cheong](2001/cheong/cheong.c) ([README.md](2001/cheong/README.md))

thanks-for-fixes.md

@@ -1084,19 +1084,24 @@ it will at least run the supplementary program as a 64-bit program directly.
 
 Cody fixed this to compile with clang but according to the author this will not
 work without i386 linux. It generates i386 32-bit code (not bytecode) but
-unfortunately it will not work without i386 linux. Cody fixed an earlier
+unfortunately it will not work without i386 linux. See [bugs.md](/bugs.md).
-segfault so that it can at least now open the file and he also changed some of
+
-the macros used to what they translate to but mostly it was kept the same.
+Cody fixed an earlier segfault so that it can at least now open the file should
+you have an i386 linux machine (it can open it in other platforms too, of
+course, but it won't work). This involved pointer updates and also changing an
+`int` to a `FILE *`. Where possible he left the macros as the same but in the
+case of `main()` this was not possible.
+
 Yusuke added another change (see below) to make it even more portable across
 compilers besides what Cody did.
 
-
+Cody entirely fixed the [supplementary
-Cody also fixed the [supplementary
 bellard.otccex.c](2001/bellard/bellard.otccex.c) so it does not segfault and
-works as well. The main problem was that some ints were being used as pointers.
+works as well (it did not work at all). The main problem was that some ints were
-This includes, for example, an int used as a `char *`, an int used as a function
+being used as pointers.  This includes, for example, an int used as a `char *`,
-pointer and an int to access `argv` as well as there being invalid access to
+an int used as a function pointer and an int to access `argv` as well as there
-`argv`. He updated the Makefile so that this program will compile by default.
+being invalid access to `argv`. He updated the Makefile so that this program
+will compile by default.
 
 Also the Fibonacci sequence (`fib()`) will overflow at `n > 48` so this is
 checked prior to running the function just like the author did for the factorial