Closed Bug 133415 Opened 23 years ago Closed 23 years ago

Crash in cert manager when switching fips mode.

Categories

(NSS :: Libraries, defect, P1)

x86
Windows 2000
defect

Tracking

(Not tracked)

VERIFIED WORKSFORME

People

(Reporter: ssaux, Assigned: rrelyea)

References

Details

To reproduce: Create a new profile. Import a certificates (e.g., sectest) from p12. You now have the certmanager window opened, showing the certificates you imported. Hide this window (e.g., with the prefs window). Open device manager. Switch to fips mode. click on the certmanager window (this will bring it to the front, and it will redraw, going to the db in the process). crash. talkback id: 4472076
Blocks: 133416
From the debugger: PK11_ReferenceSlot(PK11SlotInfoStr * 0xcdcdcdcd) line 459 + 3 bytes PK11_GetInternalKeySlot() line 2206 + 43 bytes PK11_IsLoggedIn(PK11SlotInfoStr * 0x053cb0b0, void * 0x00000000) line 2237 + 5 bytes PK11_GetAllTokens(unsigned long 0x00000001, int 0x00000000, int 0x00000001, void * 0x00000000) line 2404 + 22 bytes PK11_GetBestSlotMultiple(unsigned long * 0x0012d950, int 0x00000001, void * 0x00000000) line 2473 + 19 bytes PK11_GetBestSlot(unsigned long 0x00000001, void * 0x00000000) line 2529 + 15 bytes PK11_VerifyRecover(SECKEYPublicKeyStr * 0x054590c8, SECItemStr * 0x0012da24, SECItemStr * 0x0012d998, void * 0x00000000) line 3004 + 13 bytes DecryptSigBlock(int * 0x0012d9d8, unsigned char * 0x05457804, SECKEYPublicKeyStr * 0x054590c8, SECItemStr * 0x0012da24, char * 0x00000000) line 75 + 21 bytes VFY_CreateContext(SECKEYPublicKeyStr * 0x054590c8, SECItemStr * 0x0012da24, int 0x00000013, void * 0x00000000) line 223 + 28 bytes VFY_VerifyData(unsigned char * 0x053a8958, int 0x000002f7, SECKEYPublicKeyStr * 0x054590c8, SECItemStr * 0x0012da24, int 0x00000013, void * 0x00000000) line 435 + 21 bytes CERT_VerifySignedData(CERTSignedDataStr * 0x053a6f7c, CERTCertificateStr * 0x044fdf68, __int64 0x00039d0f3fccd230, void * 0x00000000) line 123 + 35 bytes CERT_VerifyCertChain(NSSTrustDomainStr * 0x043e6100, CERTCertificateStr * 0x053a6f70, int 0x00000001, int 0x00000000, __int64 0x00039d0f3fccd230, void * 0x00000000, CERTVerifyLogStr * 0x00000000) line 712 + 28 bytes CERT_VerifyCert(NSSTrustDomainStr * 0x043e6100, CERTCertificateStr * 0x053a6f70, int 0x00000001, int 0x00000000, __int64 0x00039d0f3fccd230, void * 0x00000000, CERTVerifyLogStr * 0x00000000) line 1138 + 37 bytes CERT_VerifyCertNow(NSSTrustDomainStr * 0x043e6100, CERTCertificateStr * 0x053a6f70, int 0x00000001, int 0x00000000, void * 0x00000000) line 1179 + 35 bytes nsNSSCertificate::GetUsageArray(char * 0x050a05f0, unsigned int * 0x0012e920, unsigned int * 0x0012e554, unsigned short * * 0x0012e560) line 1407 + 25 bytes nsNSSCertificate::GetPurposes(nsNSSCertificate * const 0x044f7570, unsigned int * 0x0012e920, unsigned short * * 0x00000000) line 2053 + 33 bytes In PK11_GetInternalKeySlot() we have mod->slots[1] which is garbage. /* get the internal key slot. FIPS has only one slot for both key slots and * default slots */ PK11SlotInfo * PK11_GetInternalKeySlot(void) { SECMODModule *mod = SECMOD_GetInternalModule(); PORT_Assert(mod != NULL); if (!mod) { PORT_SetError( SEC_ERROR_NO_MODULE ); return NULL; } return PK11_ReferenceSlot(mod->isFIPS ? mod->slots[0] : mod->slots[1]); }
This must be a regression. The bug existed a long time ago, and it was fixed a long time ago. What changes have been made to the module loading code?
Assigned the bug to Bob. The tip of the Mozilla client is using NSS 3.4 RC1 right now.
Assignee: wtc → relyea
Priority: -- → P1
Target Milestone: --- → 3.4.1
Changed the QA contact to Bishakha.
QA Contact: sonja.mirtitsch → bishakhabanerjee
Target Milestone: 3.4.1 → 3.5
John, could you try to reproduce this bug with either Mozilla 1.0 RC2 or Netscape/7.0b1? I am just wondering if this bug is actually a duplicate of bug 142659 that Bob recently fixed. Thanks.
QA Contact: bishakhabanerjee → junruh
I still get a crash in nss3.dll with this build. Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0rc2) Gecko/20020510 http://climate/reports/SingleIncidentInfo.cfm?dynamicBBID=6355771
This fixes for this bug wouldn't show up in a daily build before 0515 or 0516, and I don't believe is in the branch. bob
John, are you still seeing a crash? I don't with my builds, but I'm using tip NSS. bob
Works for me with the 6/6 branch build - no crash.
Status: NEW → RESOLVED
Closed: 23 years ago
Resolution: --- → WORKSFORME
Verified. Following the original steps exactly, I do not see a crash.
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.