New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Bleadperl v5.19.3-361-gde935cc breaks YAPPO/Data-Model-0.00008.tar.gz #13254
Comments
From @andkgit bisect commit de935cc Allow 64-bit array and stack offsets in entersub & goto diagnostics http://www.cpantesters.org/cpan/report/039a5aec-1c0b-11e3-b738-32d47e66b2dc Failing reports come only from non-threaded perls. Threaded perls pass perl -V Summary of my perl5 (revision 5 version 19 subversion 4) configuration: Characteristics of this binary (from libperl): -- |
From @jkeenanOn Fri Sep 13 00:59:28 2013, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:
The full commit message from the above was: ########## Allow 64-bit array and stack offsets in entersub & goto ######### Shouldn't a plan have been developed to have *somebody* with enough And where was its need "to be done eventually anyway" discussed? Thank you very much.
|
The RT System itself - Status changed from 'new' to 'open' |
From @cpansproutOn Sat Sep 14 05:36:36 2013, jkeenan wrote:
This would be at least 30GB, possibly more, in this case. Most cases of
Ticket #72784. If I’m working on an area of perl and I see that large integers are I haven’t looked into this failure yet, but I suspect this is a false -- Father Chrysostomos |
From @tseeOn 09/14/2013 04:12 PM, Father Chrysostomos via RT wrote:
For the record, if anybody finds a case that really, really requires It doesn't come for free in terms of my time. I'd have to rig an --Steffen |
From victor@vsespb.ruFYI there is Amazon EC2 hosting with 30Gb instances for ~ $1/hour (+ some 2013/9/15 Steffen Mueller <smueller@cpan.org>
|
From @tonycozOn Sat, Sep 14, 2013 at 11:34:01PM +0200, Steffen Mueller wrote:
The new dromedary has 96GB: [tonyc@dromedary-001 ~]$ free Tony |
From @cpansproutOn Sun Sep 15 04:24:42 2013, tonyc wrote:
That’s useful to know. The last time I tried to use that much ram I -- Father Chrysostomos |
From @tseeOn 09/15/2013 01:24 PM, Tony Cook wrote:
Ui, nice. Dennis++ for getting some of our production hardware for this. :) --Steffen |
From @cpansproutOn Fri Sep 13 00:59:28 2013, andreas.koenig.7os6VVqR@franz.ak.mind.de wrote:
This appears to be a hash randomisation issue, though I have not yet This little script: use Data::MessagePack; gives me different output each time (from 4 runs): $VAR1 = With PERL_HASH_SEED=0 the differences go away (for the one-liner, but With 5.18.1, the output is always the same. This is very strange.... -- Father Chrysostomos |
From @cpansproutOn Mon Sep 16 23:28:19 2013, sprout wrote:
The randomisation appears to be the results of arguments (that I just know that the attached patch fixes it, which makes absolutely no I have to add the I32 cast in at least those two places for the bug to Hmm, could this have something to do with mg_size returning I32, -- Father Chrysostomos |
From @cpansproutInline Patchdiff --git a/pp_hot.c b/pp_hot.c
index 1155328..469a9ee 100644
--- a/pp_hot.c
+++ b/pp_hot.c
@@ -2717,12 +2717,12 @@ try_autoload:
AV * const av = GvAV(PL_defgv);
const SSize_t items = AvFILL(av) + 1;
- if (items) {
+ if ((I32)items) {
SSize_t i = 0;
const bool m = cBOOL(SvRMAGICAL(av));
/* Mark is at the end of the stack. */
EXTEND(SP, items);
- for (; i < items; ++i)
+ for (; i < (I32)items; ++i)
{
SV *sv;
if (m) { |
From @cpansproutOn Thu Sep 19 06:29:41 2013, sprout wrote:
That was just a guess based on the fact that the cast in if((I32)items)
I began to think after writing that that it smelt a bit like an And now it appears to be exactly that: Data::MessagePack’s enc_t enc; // setup configuration svp = hv_fetchs(hv, "prefer_integer", FALSE); svp = hv_fetchs(hv, "canonical", FALSE); There are no elses here to initialise enc.prefer_int and enc.canonical So I have submitted a report to -- Father Chrysostomos |
From @cpansproutOn Thu Sep 19 13:16:59 2013, sprout wrote:
Data::Model is also at fault, in expecting canonical order without This is not a new bug, as it has failed intermittently in the past. So -- Father Chrysostomos |
@cpansprout - Status changed from 'open' to 'rejected' |
From @cpansproutOn Thu Sep 19 17:48:18 2013, sprout wrote:
I put the dot inside the angle brackets by mistake. Here is a clickable https://rt.cpan.org/Ticket/Display.html?id=76404
-- Father Chrysostomos |
Migrated from rt.perl.org#119771 (status was 'rejected')
Searchable as RT119771$
The text was updated successfully, but these errors were encountered: