No change is proposed in this patch for PORT_NewArena() because this function calls PORT_ZAlloc() and in PORT_ZAlloc(), the needed change is proposed instead. That is, on getting a memory error, the error is being set to SEC_ERROR_NO_MEMORY in PORT_ZAlloc(). Note You need to log in before you can comment on or make changes to this bug. Implemented suggestions made in Comment #9 Patch for nsprstub.c remains same as in the previous patch(272654). Source
[email protected] NewAccount | Log In or Remember [x] | Forgot Password Login: [x] Home | New | Browse | Search | [help] | Reports | Product Dashboard Last Comment Bug338688 - Member frankmorgner commented Dec 11, 2014 I guess so, reopen if not frankmorgner closed this Dec 11, 2014 vletoux referenced this issue Mar 27, 2015 Closed Memory corruption with firefox (reopen This philosophy is followed in case of other functions too. nss/lib/util/nssilock.c is different from the earlier patch.
Authentication with the smartcard is succesful and the browser prompts correctly for the certificate therein stored; however, as soon as the certificate is shown it seems that the memory gets corrupted nss/lib/ckfw/nsprstub.c remains same as in the earlier patch(272463). This is on a ZTE Blade running CyanogenMod 7.2 (Android 2.3).
The reference for the bug on fedora bugzilla is https://bugzilla.redhat.com/show_bug.cgi?id=1089476 — Reply to this email directly or view it on GitHub #232. The patch to nsprstub.c looks good. This happens both with the distribution package AND with a recompiled version of the latest git checkout. pz_NewLock() looks like a bug after your else clause, lock->ltype shouldn't be executed when lock == NULL.
Douglas E. Comment 7 Biswatosh Chakraborty 2007-07-17 08:16:40 PDT Created attachment 272652 [details] [diff] [review] For review. The change is done because the second argument to Vtrace() tries to access a member of a structure with the pointer to the structure. his comment is here This tries to incorporate the suggestions made on the previous patch (in Comment #5 and #6).
Comment 3 Biswatosh Chakraborty 2007-07-06 04:12:12 PDT Created attachment 271208 [details] [diff] [review] Not for review. Iplantotesttwothings: 1.usetheolderslowmachineide-14toverify(ide-14wassucessfulwith build5a) 2.introducetheLBintothepicture,asthisisthefirsttimewedroppedthe LB,andgoagainsttheserverdirectly. 09/Oct/2008:00:10:07]info(20388):CORE3274:successfulserverstartup [09/Oct/2008:02:59:23]failure(20388):HTTP3068:Errorreceivingrequest from 10.4.118.215(SEC_ERROR_NO_MEMORY:Outofmemory.) [09/Oct/2008:02:59:23]failure(20388):HTTP3068:Errorreceivingrequest from 10.4.118.215(SEC_ERROR_NO_MEMORY:Outofmemory.) [09/Oct/2008:02:59:23]failure(20388):HTTP3068:Errorreceivingrequest from 10.4.118.215(SEC_ERROR_NO_MEMORY:Outofmemory.) [09/Oct/2008:02:59:23]failure(20388):HTTP3068:Errorreceivingrequest from 10.4.118.215(SEC_ERROR_NO_MEMORY:Outofmemory.) [09/Oct/2008:02:59:24]warning(20388):CORE3283:stderr:Exception java.lang.O utOfMemoryError:requested32756bytesforChunkPool::allocate.Outofswap spa ce? [09/Oct/2008:02:59:26]info(20284):CORE3050:Spawningreplacementfor worker process20388 [Issue 3858] New - idp This worked (as far as I'm aware) until the 20.0 Beta update. Same as previous attachment but with some indendation changes.
It sets errors to SEC_ERROR_NO_MEMORY during memory failures It touches 2 files, namely (1)nss/lib/ckfw/nsprstub.c and (2) nss/lib/util/nssilock.c. this contact form I also suppose the problem might be with firefox (I shall have to try running it under gdb or possibly a strong malloc() checker), even if an attempt with its pre-beta Select the Google search icon Alternatively, visit https://google.com by typing it into the address bar. bob Comment 17 Biswatosh Chakraborty 2007-07-25 23:56:50 PDT Commiting nsprstub.c into NSS_3_11_BRANCH Checking in nsprstub.c; /cvsroot/mozilla/security/nss/lib/ckfw/nsprstub.c,v <-- nsprstub.c new revision: 22.214.171.124; previous revision: 1.6 done Commiting nssilock.c into NSS_3_11_BRANCH Checking in
Other names may be trademarks of their respective owners. (revision 20160708.bf2ac18) Powered by Oracle, Project Kenai and Cognisync Done Please Confirm No Yes "Fossies" - the Fresh Open Or I could (for this bug). All rights reserved. http://idroprofessional.com/general/sec-error-not-initialized.html Comment 1 Aaron Train [:aaronmt] 2013-02-27 06:50:59 PST *** This bug has been marked as a duplicate of bug 832942 *** Note You need to log in before you can comment
I request this patch for review. Yet the logic in the error paths for the two are quite different. Checking in nsprstub.c; /cvsroot/mozilla/security/nss/lib/ckfw/nsprstub.c,v <-- nsprstub.c new revision: 1.7; previous revision: 1.6 done Checking in nssilock.c; /cvsroot/mozilla/security/nss/lib/util/nssilock.c,v <-- nssilock.c new revision: 1.3; previous revision: 1.2 done Comment 15 Biswatosh Chakraborty 2007-07-25
Status: RESOLVED DUPLICATE of bug 832942 Whiteboard: Keywords: Product: Firefox for Android Classification: Client Software Component: General (show other bugs) Version: 20 Branch Platform: ARM Android Importance: -- normal (vote) TargetMilestone:
Comment 10 Biswatosh Chakraborty 2007-07-23 02:28:23 PDT Created attachment 273389 [details] [diff] [review] For review. Comment 8 Biswatosh Chakraborty 2007-07-17 08:21:55 PDT Created attachment 272654 [details] [diff] [review] For review. We're not affiliated or endorsed by the Mozilla Corporation but we love them just the same. The problem could be in OpenSC, Firefox or in the OpenSc card specific code for you card.
Comment 4 Biswatosh Chakraborty 2007-07-16 02:03:29 PDT Created attachment 272463 [details] [diff] [review] For review. Comment 16 Robert Relyea 2007-07-25 16:09:52 PDT Comment on attachment 273389 [details] [diff] [review] For review. Set the error code in both, and consistently fall through, or consistently return. (I suggest falling through). Check This Out You signed out in another tab or window.
This rule also applies to NSS's PZ_ functions that allocate locks and related objects. Peer does not support high-grade encryption." The local … Bookmark the permalink. Same as previous attachment but with some indendation changes. Comment 1 Nelson Bolyard (seldom reads bugmail) 2006-06-07 15:29:26 PDT Correction: I erroneously wrote: "the global SSL allocation failure counter".
Powered by the Content Parser System, copyright 2002 - 2015 James G. Ross. Same as previous attachment but with some indendation changes. (6.28 KB, patch) 2007-07-16 02:03 PDT, Biswatosh Chakraborty neil.williams: review- Details | Diff | Splinter Review For review. The global variable port_allocFailures should be atomically incremented, but that is the subject of another bug.
Modifies nssilock.c mainly. In some functions, it purposefully does not set the error to SEC_ERROR_NO_MEMORY in case of memory failure because that function is calling another function which actually faces this shortage of memory You signed in with another tab or window. bash-3.00#!909 grep-nstartup/export/home/fan/webserver7/https-s1/logs/errors 7:[08/Oct/2008:18:42:56]info(20285):CORE3274:successfulserverstartup 20:[08/Oct/2008:21:28:54]info(20361):CORE3274:successfulserverstartup 29:[09/Oct/2008:00:10:07]info(20388):CORE3274:successfulserverstartup 41:[09/Oct/2008:03:00:32]info(20418):CORE3274:successfulserverstartup 52:[09/Oct/2008:05:38:16]info(20450):CORE3274:successfulserverstartup bash-3.00# TheWS7errorlogiscapturedat: am-t2-06:/export/home/fan/webserver7/https-s1/logs/errorscases/3858/errors.4 Eachtime,itprinted.SEC_ERROR_NO_MEMORYiserrormessageinNSSlibrary.
In particular, I found that PORT_NewArena and PZ_NewLock do not. In the first file, it does (a)++port_allocFailures; and (b)PORT_SetError(SEC_ERROR_NO_MEMORY) in many places. What is the target of this bug? 3.12?