CALL-STACK overflow, stack size = 8712192.. upgrade FAILED for mkcl from REQUIRE using method 'load-asdf-lisp'load-asdf-fasl you can retry just that test with: ASDF_UPGRADE_TEST_TAGS="REQUIRE" ASDF_UPGRADE_TEST_METHODS="'load-asdf-lisp'load-asdf-fasl" ./test/\ run-tests.sh -u mkcl or more interactively (and maybe with rlwrap or in emacs), start with: mkcl -norc then copy/paste: (load "test/script-support.lisp") (asdf-test::da) (test-upgrade 'load-asdf-lisp'load-asdf-fasl "REQUIRE")
On Mon, Apr 21, 2014 at 4:46 PM, Robert P. Goldman rpgoldman@sift.infowrote:
CALL-STACK overflow, stack size = 8712192.. upgrade FAILED for mkcl from REQUIRE using method 'load-asdf-lisp'load-asdf-fasl you can retry just that test with: ASDF_UPGRADE_TEST_TAGS="REQUIRE" ASDF_UPGRADE_TEST_METHODS="'load-asdf-lisp'load-asdf-fasl" ./test/\ run-tests.sh -u mkcl or more interactively (and maybe with rlwrap or in emacs), start with: mkcl -norc then copy/paste: (load "test/script-support.lisp") (asdf-test::da) (test-upgrade 'load-asdf-lisp'load-asdf-fasl "REQUIRE")
The whole MKCL process has hit the OS imposed limit on the size of the stack. A call to "ulimit -s" at the shell prompt should give you a number that when multiplied by 1024 should be very close to the "stack size" claimed here above.
Exhausting 8 Meg of stack is usually sign of some form of infinite recursion in the program so I would not even advise to attempt to grow the stack...
I'll try to investigate further.
Jean-Claude Beaudoin wrote:
On Mon, Apr 21, 2014 at 4:46 PM, Robert P. Goldman <rpgoldman@sift.info mailto:rpgoldman@sift.info> wrote:
CALL-STACK overflow, stack size = 8712192.. upgrade FAILED for mkcl from REQUIRE using method 'load-asdf-lisp'load-asdf-fasl you can retry just that test with: ASDF_UPGRADE_TEST_TAGS="REQUIRE" ASDF_UPGRADE_TEST_METHODS="'load-asdf-lisp'load-asdf-fasl" ./test/\ run-tests.sh -u mkcl or more interactively (and maybe with rlwrap or in emacs), start with: mkcl -norc then copy/paste: (load "test/script-support.lisp") (asdf-test::da) (test-upgrade 'load-asdf-lisp'load-asdf-fasl "REQUIRE")The whole MKCL process has hit the OS imposed limit on the size of the stack. A call to "ulimit -s" at the shell prompt should give you a number that when multiplied by 1024 should be very close to the "stack size" claimed here above.
Exhausting 8 Meg of stack is usually sign of some form of infinite recursion in the program so I would not even advise to attempt to grow the stack...
I'll try to investigate further.
OK, this is what happens if I go to the ASDF git working copy and do
make test-upgrade l=mkcl
or, equivalently, go to asdf/test/ and do
./run-tests.sh -u mkcl
Hope that will be enough to be easily repeatable by you.
best, r
On Mon, Apr 21, 2014 at 10:53 PM, Robert P. Goldman rpgoldman@sift.infowrote:
OK, this is what happens if I go to the ASDF git working copy and do
make test-upgrade l=mkcl
or, equivalently, go to asdf/test/ and do
./run-tests.sh -u mkcl
Hope that will be enough to be easily repeatable by you.
I had to do this instead:
./test/run-tests.sh -u mkcl
because the run-test.sh shell script seems to be picky about the relative directory structure it runs in.
The script tries 3 other successful upgrade methods before blowing up on the 4th attempt. At least with the changes I just made there is no call stack exhaustion anymore, just a normal error being signaled which ends up in the printing of a stack backtrace.
Working on it...
On Tue, Apr 22, 2014 at 1:26 AM, Jean-Claude Beaudoin < jean.claude.beaudoin@gmail.com> wrote:
I had to do this instead:
./test/run-tests.sh -u mkcl
... Working on it...
Could you please apply the content of this tiny attached patch and tell me if it makes the problem go away?
Thanks,
jcb
On Tue, Apr 22, 2014 at 4:14 AM, Jean-Claude Beaudoin jean.claude.beaudoin@gmail.com wrote:
On Tue, Apr 22, 2014 at 1:26 AM, Jean-Claude Beaudoin jean.claude.beaudoin@gmail.com wrote:
I had to do this instead:
./test/run-tests.sh -u mkcl
... Working on it...
Could you please apply the content of this tiny attached patch and tell me if it makes the problem go away?
Works for me.
pushed as 3.1.0.117.
—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org La superstition porte malheur. — Paul Carvel
Faré wrote:
On Tue, Apr 22, 2014 at 4:14 AM, Jean-Claude Beaudoin jean.claude.beaudoin@gmail.com wrote:
On Tue, Apr 22, 2014 at 1:26 AM, Jean-Claude Beaudoin jean.claude.beaudoin@gmail.com wrote:
I had to do this instead:
./test/run-tests.sh -u mkcl
... Working on it...
Could you please apply the content of this tiny attached patch and tell me if it makes the problem go away?
Works for me.
pushed as 3.1.0.117.
MKCL now passes upgrade tests for me, too!
Calling it good....
-
Faré -
Jean-Claude Beaudoin -
Robert P. Goldman