bencoh | yay :) | 00:01 |
---|---|---|
*** esoare has quit IRC | 00:22 | |
*** Pali has quit IRC | 01:38 | |
*** RedW has quit IRC | 02:01 | |
*** RedW has joined #maemo-ssu | 02:06 | |
*** arcean has quit IRC | 02:17 | |
*** rd_ has joined #maemo-ssu | 02:29 | |
xes | freemangordon: well done! | 02:31 |
*** _rd has quit IRC | 02:33 | |
*** M4rtinK has quit IRC | 02:38 | |
*** dafox has quit IRC | 03:47 | |
*** futpib has quit IRC | 04:17 | |
*** nox- has quit IRC | 06:14 | |
*** rd_ has quit IRC | 09:15 | |
*** arcean has joined #maemo-ssu | 10:27 | |
*** RzR has quit IRC | 10:45 | |
*** RzR has joined #maemo-ssu | 10:46 | |
*** rd_ has joined #maemo-ssu | 11:02 | |
*** sparetire_ has quit IRC | 11:13 | |
freemangordon | merlin1991: I think it is relatively safe o include openssl in the next -testing | 11:37 |
freemangordon | *to | 11:37 |
freemangordon | if there is something broken, I doubt we'll catch it while the package is in -devel. so far I tested everything I can think about and there seems to be no problem - https, email (imap tls), ssh, supl. | 11:39 |
bencoh | maybe we should nm all the binaries and grep for affected symbols | 11:40 |
bencoh | might help catching potential issues | 11:40 |
bencoh | but the main problem here is things will probably "work" but we might end up with invalid reads/writes at some point | 11:41 |
freemangordon | bencoh: symbols should be there, it is supposed to be the same ABI | 11:48 |
bencoh | yeah but some structures differ | 11:49 |
freemangordon | but you won;t catch that with nm | 11:49 |
bencoh | some of which might be instantiated from the app and passed to functions | 11:49 |
bencoh | no, but you can grep on the functions which are passed those struct/struct* | 11:49 |
freemangordon | I doubt there is a transparent structures to be used in applications | 11:49 |
freemangordon | *are | 11:49 |
freemangordon | if an application uses such, then I'll call it buggy | 11:50 |
freemangordon | even if it malloc's it | 11:50 |
bencoh | there are structures defined in headers | 11:50 |
freemangordon | bencoh: but feel free to do ^^^ if you think it is possible | 11:51 |
freemangordon | so? | 11:51 |
bencoh | they're not private/opaque :) | 11:51 |
bencoh | so, I dont know | 11:51 |
bencoh | hm, headers look bloated in openssl :D | 11:52 |
freemangordon | :nod: | 11:52 |
kerio | let's just move to libressl | 11:55 |
kerio | we only need to recompile everything | 11:55 |
bencoh | :/ | 11:56 |
bencoh | there are things we cant rebuild :( | 11:56 |
kerio | :( | 11:57 |
bencoh | most most datatype changes reported in http://upstream.rosalinux.ru/versions/openssl.html say "99 symbols affected" (basically every symbole, more or less) | 12:00 |
kerio | wut | 12:02 |
bencoh | that means nm+grep wont work | 12:03 |
freemangordon | bencoh: look at the correct version, 0.9.8 that is | 12:03 |
kerio | yeah i mean | 12:05 |
kerio | we're going 0.9.8n to 0.9.8zf | 12:05 |
freemangordon | :nod: | 12:06 |
*** Pali has joined #maemo-ssu | 12:06 | |
bencoh | freemangordon: yeah that's what I'm checking :) | 12:07 |
bencoh | looks like za reverts datatype issues introduced in y | 12:07 |
bencoh | so we're left with issues introduced in p | 12:08 |
bencoh | which might be solved just by keeping the old PQ_64BIT behaviour | 12:09 |
bencoh | I wouldnt be surprised if debian did that for us | 12:09 |
kerio | i like how openssl has shit like "is_probably_safari" | 12:10 |
bencoh | :)) | 12:11 |
bencoh | kerio: then what about PQ_64BIT_IS_BIGNUM (bignum was 20bytes ... 160bits) ? ;p | 12:11 |
freemangordon | bencoh: could you check the patches? | 12:12 |
freemangordon | https://github.com/community-ssu/openssl/tree/0.9.8zf-1%2Bmaemo1%2B0m5%2B0cssu0/debian/patches | 12:13 |
freemangordon | hmm, debian has patches up to 0.9.8o :( | 12:15 |
freemangordon | I guess it is because of the p breaking ABI | 12:15 |
bencoh | is this patch what I think ? https://github.com/community-ssu/openssl/blob/master/debian/patches/12_valgrind.patch | 12:15 |
bencoh | reminds me of an (in)famous debian patch | 12:16 |
bencoh | I think someone had a look at it before, but ... | 12:16 |
bencoh | freemangordon: hm probably yeah | 12:16 |
bencoh | they keep backporting changes in squeeze-lts | 12:17 |
bencoh | why not use squeeze-lts as upstream for openssl ? | 12:17 |
freemangordon | http://anonscm.debian.org/viewvc/pkg-openssl/openssl/tags/0.9.8o-4squeeze15/debian/patches/ | 12:17 |
freemangordon | because the newest patch there is 9 months old | 12:18 |
bencoh | https://packages.qa.debian.org/o/openssl/news/20150320T214900Z.html | 12:18 |
freemangordon | hmm, where did you get that from? | 12:19 |
freemangordon | I mean - where is the source repo? | 12:19 |
bencoh | dunno | 12:19 |
bencoh | I didnt get it - yet | 12:19 |
bencoh | and ... google cant find the packages | 12:20 |
bencoh | http://ftp.de.debian.org/debian/pool/main/o/openssl/openssl_0.9.8o-4squeeze20.debian.tar.gz | 12:21 |
bencoh | google hasnt indexed it | 12:21 |
bencoh | (yet) | 12:21 |
freemangordon | we'd rather fix the ABI breakage (if any) | 12:23 |
bencoh | the valgrind.patch is still in debian, so I guess (hope ? :D) it's not the infamous 2008ish debianssl bg | 12:23 |
bencoh | bug | 12:24 |
bencoh | freemangordon: there our best bet is probably defining PQ_64BIT_IS_BIGNUM to 1 and PQ_64BIT_IS_INTEGER to 0 | 12:24 |
bencoh | (and maybe the other two PQ_64BIT constants) | 12:24 |
freemangordon | bencoh: look at pqueue/pq_compat.h | 12:27 |
bencoh | yeah | 12:29 |
freemangordon | and it seems defined in SB | 12:29 |
freemangordon | ~/workspace/cssu/openssl/git/openssl-0.9.8zf] > grep -r THIRTY * | 12:29 |
bencoh | oh ? | 12:29 |
freemangordon | ... | 12:29 |
freemangordon | crypto/opensslconf.h:#define THIRTY_TWO_BIT | 12:29 |
bencoh | :D | 12:29 |
bencoh | nice, we might be safe then | 12:30 |
freemangordon | I'll put some garbage in that header, to make sure this is really defained | 12:30 |
bencoh | hm we need to make sure it's used everywhere | 12:31 |
freemangordon | ../../include/openssl/pq_compat.h:113:2: error: #error "# if (defined(THIRTY_TWO_BIT) && !defined(BN_LLONG)) || defined(SIXTEEN_BIT) || defined(EIGHT_BIT) *IS NOT* defined, we are busted :(" | 12:35 |
freemangordon | make[3]: *** [pqueue.o] Error 1 | 12:35 |
bencoh | okay, that's the only place PQ_64BIT is defined, so we should be safe | 12:35 |
bencoh | oh ? | 12:35 |
bencoh | hmm right, we're 32b but we have long long support on arm | 12:36 |
freemangordon | yeah | 12:38 |
freemangordon | that one should be fixed. I guess | 12:38 |
freemangordon | however, I gtg, will look at it later | 12:38 |
bencoh | that means we give up on long long optim ... well, if those didnt exist before 0.9.8p, we wont miss anything | 12:39 |
freemangordon | well, we can try to wrap | 12:41 |
freemangordon | those are just a couple of functions anyway | 12:42 |
bencoh | yeah | 12:43 |
bencoh | funny thing is ... I cant find those changes in openssl git (git log between tags) :/ | 12:43 |
bencoh | dunno what they did there | 12:43 |
freemangordon | check on github, they have "file history" | 12:44 |
freemangordon | https://github.com/openssl/openssl | 12:44 |
bencoh | ? | 12:44 |
freemangordon | https://github.com/openssl/openssl/commits/OpenSSL_0_9_8zf/crypto/pqueue/pqueue.h | 12:45 |
freemangordon | https://github.com/openssl/openssl/commit/e2ac4732cd4f0efc3dff3261d19aed0ced2c4dde | 12:45 |
freemangordon | hmm, wait, isn;t that pq shit internal? | 12:46 |
bencoh | no, it's reflected in public structures | 12:46 |
bencoh | dtls1_/ssl3_ stuff | 12:47 |
freemangordon | hmm, yeah | 12:47 |
freemangordon | https://github.com/openssl/openssl/commit/e2ac4732cd4f0efc3dff3261d19aed0ced2c4dde#diff-bfa5a0561be6e4126229a7fd57bc24e0R299 | 12:47 |
bencoh | ssl3_record_st has seq_num for instance | 12:47 |
bencoh | right ;) | 12:47 |
freemangordon | ok, we can wrap it | 12:48 |
freemangordon | or maybe not | 12:48 |
bencoh | :] | 12:48 |
freemangordon | wait... do we really need to fix that? | 12:49 |
bencoh | dunno | 12:49 |
freemangordon | the new structure is smaller than the old one | 12:49 |
freemangordon | merlin1991: Pali: ^^^ | 12:50 |
*** amiconn has quit IRC | 12:50 | |
freemangordon | maybe someone should ask on #openssl | 12:52 |
freemangordon | or whateer the channel is | 12:52 |
*** amiconn has joined #maemo-ssu | 12:52 | |
* freemangordon is afk | 12:53 | |
Pali | I really do not know :-( | 12:53 |
freemangordon | Pali: noone knows, that is why we investigate :) | 12:53 |
freemangordon | bencoh: could you check what sizeof(those 2 structs) gives with old and with new openssl? | 12:58 |
bencoh | I'm checking diffs between (current cssu) openssl-0.9.8n and zf, and ... | 12:58 |
bencoh | diff -w -u crypto/pqueue/pq_compat.h ../openssl-cssu/openssl-0.9.8zf/crypto/pqueue/pq_compat.h | 12:58 |
bencoh | returns nothing | 12:58 |
bencoh | I feel like I missed something at some point | 12:59 |
bencoh | or upstream.rosalinux.ru says shit | 12:59 |
freemangordon | keep it mind that it might say shit, becuse of the missing arch specific stuff | 13:01 |
bencoh | true | 13:01 |
bencoh | it might just be that | 13:01 |
freemangordon | that is why the real test should be comparing the sizeof structs, just to be sure | 13:01 |
freemangordon | but I don;t have time now | 13:02 |
freemangordon | buy, bbl | 13:02 |
bencoh | should I just compile a oneliner against the headers ? | 13:02 |
*** M4rtinK has joined #maemo-ssu | 13:28 | |
*** futpib has joined #maemo-ssu | 13:41 | |
bencoh | I should have disabled the checks in debian/rules ... it's taking ages | 13:55 |
*** trx has joined #maemo-ssu | 14:02 | |
*** rd_ has quit IRC | 14:08 | |
bencoh | freemangordon: http://pastebin.notk.org/pastebin.php?show=m6cc01d7d | 14:10 |
bencoh | looks good | 14:10 |
*** rd_ has joined #maemo-ssu | 14:17 | |
Pali | why is git read-tree so sloooow? | 14:34 |
Pali | probably it removes all inodes information from .git/index... | 14:35 |
*** rd_ has quit IRC | 14:58 | |
*** rd_ has joined #maemo-ssu | 15:00 | |
bencoh | http://pastebin.notk.org/pastebin.php?show=m597f644e | 15:03 |
Pali | only two structs? | 15:22 |
Pali | this is not too much | 15:23 |
*** arcean has quit IRC | 15:28 | |
*** arcean has joined #maemo-ssu | 15:38 | |
*** arcean has quit IRC | 15:51 | |
bencoh | some structs might be missing in this test, but at least it covers the pq64 thing, and a few others | 16:08 |
freemangordon | bencoh: yes, looks good | 17:07 |
freemangordon | except that x509_cinf_st thingie | 17:08 |
bencoh | yeah, dunno where that comes from | 17:08 |
freemangordon | bencoh: does headers differ? | 17:12 |
freemangordon | it is crypto/x509/x509.h | 17:12 |
bencoh | hmm right, looks like they added ASN1_ENCODING enc | 17:15 |
bencoh | couldnt see it in git log :) | 17:15 |
freemangordon | bencoh: https://github.com/openssl/openssl/commit/6cb5746b65a53207bc794121a0c78603bff8ea30 | 17:16 |
bencoh | and hm_fragment_st has unsigned char *reassembly | 17:16 |
freemangordon | bencoh: anyway X509_CINF structure should not be allocated with malloc, there is X509_CINF *X509_CINF_new(); function | 17:18 |
bencoh | hmm, so we should be fine | 17:18 |
bencoh | hm_fragment doesnt seem to be used in any public function (?!) | 17:19 |
freemangordon | I don't see it used at all, weird | 17:22 |
freemangordon | oh | 17:22 |
freemangordon | it is hw_fragment | 17:22 |
freemangordon | yep, seems like internal struct | 17:23 |
freemangordon | bencoh: seems the ABI is the same | 17:23 |
bencoh | hm_* not hw_* | 17:24 |
freemangordon | yeah, a typo | 17:24 |
bencoh | but ... yeah, everything's fine in the end | 17:24 |
freemangordon | :) | 17:24 |
bencoh | I guess I just worried too much :) | 17:24 |
freemangordon | no, it is ok, better be on the safe side with such a critical library | 17:25 |
*** futpib has quit IRC | 18:06 | |
*** rd_ has quit IRC | 18:29 | |
*** rd_ has joined #maemo-ssu | 18:29 | |
*** xes_ has joined #maemo-ssu | 18:31 | |
*** xes has quit IRC | 18:33 | |
*** LauRoman has joined #maemo-ssu | 18:46 | |
kerio | freemangordon: do i yolo upgrade | 19:13 |
kerio | are there some reported breakages | 19:14 |
freemangordon | kerio: nothing broken so far | 19:15 |
kerio | alright i found a packaging problem already | 19:15 |
kerio | openssl doesn't depend on libssl0.9.8 >= 0.9.8zf | 19:15 |
freemangordon | I even managed to connect to FB with QWebView | 19:15 |
freemangordon | sure, why it should depend on that particular varion? | 19:15 |
freemangordon | *version | 19:16 |
kerio | because... it's openssl 0.9.8zf | 19:16 |
freemangordon | but, ABI is 0.9.8 | 19:16 |
kerio | k | 19:16 |
freemangordon | not 0.9.8zf | 19:16 |
kerio | freemangordon: there's a new busybox-power apparently | 19:17 |
freemangordon | kerio: I have only 2 hands, still | 19:18 |
freemangordon | and 24hours per day | 19:18 |
freemangordon | deffinitely, openssl is of higher priority | 19:18 |
kerio | freemangordon: http://www.blogcdn.com/www.engadget.com/media/2008/05/5-29-08-carpet-keyboard.jpg | 19:18 |
freemangordon | those are legs, not hands | 19:19 |
kerio | mh, who uploaded libc6 +cssu3? | 19:21 |
freemangordon | merlin1991, afaik | 19:22 |
kerio | why was a non-thumb libc6 in your repo | 19:22 |
merlin1991 | kerio: +cssu3 is in -devel not in -thumb | 19:23 |
kerio | +cssu2 is in -thumb | 19:23 |
*** LauRoman has quit IRC | 19:58 | |
*** rd_ has quit IRC | 20:03 | |
freemangordon | kerio: because thumb-compiled seemed to bring ptoblems | 20:09 |
kerio | ;-; | 20:09 |
freemangordon | so I increased the version, ARM-compiled it and put it in the repo | 20:09 |
*** LauRoman has joined #maemo-ssu | 20:11 | |
kerio | k | 20:13 |
*** sparetire_ has joined #maemo-ssu | 20:32 | |
*** rd_ has joined #maemo-ssu | 20:39 | |
*** arcean has joined #maemo-ssu | 20:41 | |
freemangordon | Pali: did you backout linux-n900 on github? | 20:43 |
freemangordon | *backup | 20:43 |
*** M4rtinK has quit IRC | 20:45 | |
*** futpib has joined #maemo-ssu | 20:47 | |
Pali | freemangordon: week ago... https://github.com/pali/linux-n900 | 20:53 |
freemangordon | Pali: ok | 20:55 |
freemangordon | kerio: http://maemo.merlin1991.at/cssu/community-thumb/pool/free/b/busybox-power/ | 20:56 |
kerio | yey | 20:56 |
*** LauRoman has quit IRC | 21:01 | |
*** freemangordon has quit IRC | 21:10 | |
*** freemangordon has joined #maemo-ssu | 21:11 | |
*** M4rtinK has joined #maemo-ssu | 21:25 | |
*** M4rtinK has quit IRC | 21:35 | |
*** M4rtinK has joined #maemo-ssu | 21:38 | |
*** M4rtinK has quit IRC | 22:10 | |
*** arcean has quit IRC | 22:12 | |
*** arcean has joined #maemo-ssu | 22:14 | |
*** M4rtinK has joined #maemo-ssu | 22:23 | |
*** M4rtinK has quit IRC | 22:27 | |
*** futpib has quit IRC | 22:40 | |
*** futpib has joined #maemo-ssu | 22:40 | |
*** futpib has quit IRC | 22:45 | |
*** futpib has joined #maemo-ssu | 23:04 | |
*** futpib has quit IRC | 23:09 | |
*** futpib_ has joined #maemo-ssu | 23:10 | |
*** futpib_ has quit IRC | 23:15 | |
*** dafox has joined #maemo-ssu | 23:21 |
Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!