<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
<br>
I tried building the tip. First, let me qualify, while I'm handy
with Lisp itself (and getting emacs running), I'm from more of a
MS/Devstudio background. So if I overlook something in what you've
said or something obvious it's not laziness or belligerence; but I
may generally not have gotten it. <br>
<br>
As to your question: It's theoretically possible that if I get you
a provisioning profile you could run a binary on a modern iPod; but
to do any real debugging you'd need XCode etc. . .<br>
<br>
<br>
What I just attempted:<br>
1. I just reimplemented the instructions at the link I sent you
earlier, to the same results - simulator worked, device build looped
churning ram until running out of ram.<br>
<br>
2. I pulled the tip from sourceforge git clone
git://ecls.git.sourceforge.net/gitroot/ecls/ecl and configured/built
it standalone on the mac. This compiled; but not meaningful for
iOS.<br>
<br>
3. I pulled the tip from sourceforge git clone
git://ecls.git.sourceforge.net/gitroot/ecls/ecl and ran the build.sh
script in the instructions. I did NOT run the patch (as it would
corrupt the makefile in a prior attempt). However, hit a
compile/link error<br>
<br>
Step 3 yielded the error below: Is this something trivial I can
change?<br>
<br>
<br>
/Developer/Platforms/iPhoneSimulator.platform/Developer/usr/bin/gcc-4.2
-fPIC -fno-common -DPACKAGE_NAME=\"gc\" -DPACKAGE_TARNAME=\"gc\"
-DPACKAGE_VERSION=\"7.1\" "-DPACKAGE_STRING=\"gc 7.1\""
-DPACKAGE_BUGREPORT=\<a class="moz-txt-link-rfc2396E" href="mailto:Hans.Boehm@hp.com\">"Hans.Boehm@hp.com\"</a> -DGC_VERSION_MAJOR=7
-DGC_VERSION_MINOR=1 -DPACKAGE=\"gc\" -DVERSION=\"7.1\"
-DGC_DARWIN_THREADS=1 -DTHREAD_LOCAL_ALLOC=1
-DHAS_X86_THREAD_STATE32___EAX=1 -DSTDC_HEADERS=1
-DHAVE_SYS_TYPES_H=1 -DHAVE_SYS_STAT_H=1 -DHAVE_STDLIB_H=1
-DHAVE_STRING_H=1 -DHAVE_MEMORY_H=1 -DHAVE_STRINGS_H=1
-DHAVE_INTTYPES_H=1 -DHAVE_STDINT_H=1 -DHAVE_UNISTD_H=1
-DHAVE_DLFCN_H=1 -DNO_EXECUTE_PERMISSION=1 -DALL_INTERIOR_POINTERS=1
-DGC_GCJ_SUPPORT=1 -DJAVA_FINALIZATION=1 -DATOMIC_UNCOLLECTABLE=1
-DLARGE_CONFIG=1
-I/Users/mlamari/ecl/kriyative-ecl-iphone-builder-16a9276/ecl_tip_3/src/gc/include
-fexceptions -I libatomic_ops/src -g -arch i386
-I/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneSimulator4.2.sdk/usr/include
-fmessage-length=0 -pipe -std=c99 -Wno-trigraphs -fpascal-strings
-fasm-blocks -O0 -Wreturn-type -Wunused-variable -isysroot
/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneSimulator4.2.sdk
-fexceptions -fvisibility=hidden -mmacosx-version-min=10.6 -gdwarf-2
-fobjc-abi-version=2 -D__IPHONE_OS_VERSION_MIN_REQUIRED=30000
-D_DARWIN_USE_64_BIT_INODE -DAPPLE -DIPHONE -DIPHONE_SIMULATOR
-DIPHONE_SDK_VER=420 -g -arch i386
-I/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneSimulator4.2.sdk/usr/include
-fmessage-length=0 -pipe -std=c99 -Wno-trigraphs -fpascal-strings
-fasm-blocks -O0 -Wreturn-type -Wunused-variable -isysroot
/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneSimulator4.2.sdk
-fexceptions -fvisibility=hidden -mmacosx-version-min=10.6 -gdwarf-2
-fobjc-abi-version=2 -D__IPHONE_OS_VERSION_MIN_REQUIRED=30000
-D_DARWIN_USE_64_BIT_INODE -DAPPLE -DIPHONE -DIPHONE_SIMULATOR
-DIPHONE_SDK_VER=420 -D_THREAD_SAFE -MT dyn_load.lo -MD -MP -MF
.deps/dyn_load.Tpo -c
/Users/mlamari/ecl/kriyative-ecl-iphone-builder-16a9276/ecl_tip_3/src/gc/dyn_load.c
-o dyn_load.o<br>
/Users/mlamari/ecl/kriyative-ecl-iphone-builder-16a9276/ecl_tip_3/src/gc/dyn_load.c:
In function 'GC_init_dyld':<br>
/Users/mlamari/ecl/kriyative-ecl-iphone-builder-16a9276/ecl_tip_3/src/gc/dyn_load.c:1151:
error: '_dyld_bind_fully_image_containing_address' is unavailable
(declared at
/Developer/Platforms/iPhoneSimulator.platform/Developer/SDKs/iPhoneSimulator4.2.sdk/usr/include/mach-o/dyld.h:241)<br>
{standard input}:69:non-relocatable subtraction expression,
"L_GC_no_dls$non_lazy_ptr" minus "L00000000001$pb"<br>
{standard input}:69:symbol: "L_GC_no_dls$non_lazy_ptr" can't be
undefined in a subtraction expression<br>
{standard input}:unknown:Undefined local symbol
L_GC_no_dls$non_lazy_ptr<br>
make[2]: *** [dyn_load.lo] Error 1<br>
make[1]: *** [install-recursive] Error 1<br>
make: *** [libeclgc.a] Error 2<br>
Regent:ecl_tip_3 mlamari$ <br>
<br>
<br>
On 1/30/2011 12:22 PM, Juan Jose Garcia-Ripoll wrote:
<blockquote
cite="mid:AANLkTinUWc+WkBn60gWYxWs1kFq2E7JdWuFjz4ME4xMO@mail.gmail.com"
type="cite">
<div class="gmail_quote">On Sun, Jan 30, 2011 at 9:15 AM, Matt
Lamari <span dir="ltr"><<a moz-do-not-send="true"
href="mailto:matt.lamari@gmail.com">matt.lamari@gmail.com</a>></span>
wrote:<br>
<blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt
0.8ex; border-left: 1px solid rgb(204, 204, 204);
padding-left: 1ex;">
I've built the <a moz-do-not-send="true"
href="https://github.com/kriyative/ecl-iphone-builder/tree/centaur"
target="_blank">https://github.com/kriyative/ecl-iphone-builder/tree/centaur</a><br>
centaur branch. It's quite involved; but seems to work.<br>
</blockquote>
<div><br>
</div>
<div>I just learned about the project from your email. I did not
know it existed. One thing you should notice is that it is for
an older version of ECL, 10.4, which may cause problems when
used with more recent versions. Are you following strictly the
instructions in github?</div>
<div><br>
</div>
<div>Another thing is that I have committed in today 90% of the
patches in ecl-ios-patches.txt, but there are some that I feel
unsure about and would like the author to help in deciphering
them (left a comment at his wall, since I found no other form
of contact). In particular, the only patches left out are a
manual editing of src/configure (no patches for aclocal.m4/<a
moz-do-not-send="true" href="http://configure.in">configure.in</a>),
and a file src/gmp/mpn/arm/invert_limb.asm which is not only
fixed but "upgraded" to version 3.0 of the LGPL, which I do
not want to use in ECL.</div>
<br>
Out of the remaining information that is in your email, I can
figure out very little: it may be due to using some more recent
version of ECL, it may be a bug in the 10.4 that is not present
now, ... In particular debug information is wrong: the statement
"x = cl_core.packages_to_be_created;" is not causing
init_lib_LSP to run, because it is a variable assignment --
seems that the compiler missed a few lines --.</div>
<div class="gmail_quote"><br>
</div>
<div class="gmail_quote">Other than this, if it runs perfectly in
the simulator and not in the iOS, I suspect it may be due to
some data corruption which is not obvious in the former. Or does
the compiler work differently in one of them? Does it produce
different code?</div>
<div class="gmail_quote"><br>
</div>
<div class="gmail_quote">My recommended line of action in this
case would be fix the github sources to build with a more recent
version of ECL (ideally CVS/git), then check it in the simulator
and then do a more detailed debug of the resulting binary -- can
it be done on the iPod touch too? I have one.<br>
<div><br>
</div>
<div>Juanjo</div>
</div>
<br>
-- <br>
Instituto de Física Fundamental, CSIC<br>
c/ Serrano, 113b, Madrid 28006 (Spain) <br>
<a moz-do-not-send="true"
href="http://juanjose.garciaripoll.googlepages.com"
target="_blank">http://juanjose.garciaripoll.googlepages.com</a><br>
</blockquote>
<br>
</body>
</html>