summaryrefslogtreecommitdiff
path: root/src/lib/libcrypto/x509 (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* Simplify and explain expand_addr() a bittb2021-12-281-12/+23
| | | | | | | | | | | | | | | | RFC 3779 section 2.1.2 does a decent job of explaining how IP addresses are encoded in. What's stored amounts to a prefix with all trailing zero octets omitted. If there are trailing zero bits in the last non-zero octet, bs->flags & 7 indicates how many. addr_expand() expands this to an address of length 4 or 16 depending on whether we deal with IPv4 or IPv6. Since an address can be the lower or the upper bound of a prefix or address range, expansion needs to be able to zero-fill or one-fill the unused bits/octets. No other expansion is ever used, so simplify the meaning of fill accordingly. There's no need to special case the case that there are no unused bits, the masking/filling is a noop. ok jsing
* Add a comment so I don't forget to think about input validationtb2021-12-281-1/+3
| | | | in make_IPAddressFamily()
* Convert make_IPAddressFamily to CBS/CBBtb2021-12-281-13/+26
| | | | | | | | | | | | | | | The IPAddrBlocks type, which represents the IPAddrBlocks extension, should have exactly one IPAddressFamily per AFI+SAFI combination to be delegated. make_IPAddressFamily() first builds up a search key from the afi and safi arguments and then looks for an existing IPAddressFamily with that key in the IPAddrBlocks that was passed in. It returns that if it finds it or allocates and adds a new one. This diff preserves the current behavior that the afi and *safi arguments are truncated to 2 and 1 bytes, respectively. This may change in the future. ok inoguchi jsing
* Remove two pointless NULL checks and allocationstb2021-12-281-7/+1
| | | | | | | The ASN.1 template for IPAddressFamily doesn't mark either of its two members as optional, so they are allocated by IPAddressFamily_new(). ok inoguchi jsing
* Check for trailing garbage in X509_addr_get_afi()tb2021-12-281-1/+5
| | | | | | | | | | | | Per RFC 3779 2.2.3.3, the addressFamily field contains the 2-byte AFI and an optional 1-byte SAFI. Nothing else. The optional SAFI is nowhere exposed in the API. It is used expliclty only for pretty printing. There are implicit uses in a few places, notably for sorting/comparing where trailing garbage would be erroneously taken into account. Erroring in this situation will let us avoid this in upcoming revisions. ok inoguchi jsing
* Convert X509v3_adr_get_afi() to CBStb2021-12-281-6/+21
| | | | | | | | | | | The manual byte bashing is performed more safely using this API which would have avoided the out-of-bounds read that this API had until a few years back. The API is somewhat strange in that it uses the reserved AFI 0 as an in-band error but it doesn't care about the reserved AFI 65535. ok inoguchi jsing
* zap doubled semicolontb2021-12-261-2/+2
|
* Fix some weird line wrapping and a minor KNF nittb2021-12-251-10/+6
|
* No need for assert.h in here.tb2021-12-251-2/+1
|
* drop a meaningless XXXtb2021-12-251-2/+1
|
* Use C99 initializers for v3_addr, v3_asid and v3_ct_scts[]tb2021-12-252-24/+30
| | | | | | as is done for most other X.509 v3 extension methods. discussed with jsing
* Undo commenting of OPENSSL_NO_RFC3779tb2021-12-241-4/+4
| | | | | | | | | | | The define implies that we have the RFC 3779 API and corresponding symbols publicly exposed. We don't do that since there are still concerns about its suitability and security. oss-fuzz has code depending on this define and this broke its build as tracked down by jsing. This commit gets us oss-fuzz builds back while keeping job happy since the extension pretty printing will continue to work. ok jsing
* Fix a typo in a comment and add some empty lines for readabilitytb2021-12-241-2/+6
|
* Fix some KNF issues in the RFC 3779 section that have bothered me fortb2021-12-241-54/+55
| | | | way too long.
* KNF nittb2021-12-241-2/+2
|
* Remove asserts from asid_validate_path_internal()tb2021-12-241-11/+22
| | | | | | | | | | | The first asserts ensure that things checked in the callers hold true. Turn them into error checks and set the error on the X509_STORE_CTX if it's present. Checking sk_value(..., i) with i < sk_num(...) isn't useful, particularly if that check is done via an assert. Turn one remaining assert into a NULL check. Finally, simplify the sk_num() checks in the callers. ok jsing
* Turn asserts in ASIdentifierChoice_canonize() into error checkstb2021-12-241-3/+5
| | | | | | | | | The first assert ensures that a stack that was just sorted in a stronger sense is sorted in a weak sense and the second assert ensures that the result of the canonization procedure is canonical. All callers check for error, so these asserts don't do anything useful. ok jsing
* Remove assert from extract_min_max() (again)tb2021-12-241-3/+1
| | | | | | All callers ensure that aor != NULL, so this isn't necessary. ok jsing
* Revert previous. The commit contained more than intended.tb2021-12-241-25/+14
|
* Turn asserts in ASIdentifierChoice_canonize() into error checkstb2021-12-241-12/+25
| | | | | | | | | The first assert ensure that a stack that was just sorted in a stronger sense is sorted in a weak sense and the second assert ensures that the result of the canonization procedure is canonical. All callers check for error, so these asserts don't do anything useful. ok jsing
* Remove assert from extract_min_max()tb2021-12-241-3/+1
| | | | | | All callers ensure that aor != NULL, so this isn't necessary. ok jsing
* Fix indent of a comment.tb2021-12-241-2/+2
|
* Remove asserts from addr_validate_path_internal()tb2021-12-241-9/+19
| | | | | | | | | | This is reachable from x509_verify(), but all asserts are previously checked in the caller. Turn them into error checks and make sure the error is set on the X509_STORE_CTX if present. Change some stack == NULL || sk_num(stack) == 0 checks into sk_num(stack) <= 0 which is equivalent but simpler. ok jsing
* Turn assert in X509v3_addr_canonize() into an error check.tb2021-12-241-3/+5
| | | | | | | All internal callers check the return value and future external callers will be happy not to hit an assert from the library. ok jsing
* Fully check the second strtoul() call in v2i_IPAddrBlocks()tb2021-12-231-3/+34
| | | | | | | | | This can read a value in an arbitrary base from a string that is supposed to be followed by whitespace or a colon, so it cannot be switched to strtonum(). The current checks don't allow a read past the end, but let's use the standard idiom instead. ok jsing
* Fix an arbitrary out-of-bounds stack read in v2i_IPAddrBlocks()tb2021-12-231-3/+7
| | | | | | | | | | | | | | | | | | | | | | | Switch an insufficiently checked strtoul() to strtonum(). This can be used to trigger a read of a user-controlled size from the stack. $ openssl req -new -addext 'sbgp-ipAddrBlock = IPv4:192.0.2.0/12341234' Segmentation fault (core dumped) The bogus prefix length 12341234 is fed into X509v3_addr_add_prefix() and used to read (prefixlen + 7) / 8 bytes from the stack variable 'min[16]' that ends up as 'data' in the memmove in ASN1_STRING_set(). The full fix will add length checks to X509v3_addr_add_prefix() and make_addressPrefix() and will be dealt with later. The entire X509v3_{addr,asid}_* API will need a thorough review before it can be exposed. This code is only enabled in -current and can only be reached from openssl.cnf files that contain sbgp-ipAddrBlock or from the openssl(1) command line. ok jsing
* Reinstate the licenses that were replaced with a license stubtb2021-12-182-12/+108
| | | | in OpenSSL commit d2e9e320.
* Include evp_locl.h where it will be needed once most structs fromtb2021-12-123-3/+6
| | | | | | evp.h will be moved to evp_locl.h in an upcoming bump. ok inoguchi
* Convert {i2d,d2i}_{,EC_,DSA_,RSA_}PUBKEY{,_bio,_fp}() to templated ASN1jsing2021-12-031-97/+1
| | | | | | | These functions previously used the old ASN1_{d2i,i2d}_{bio,fp}() interfaces. ok inoguchi@ tb@
* Bugfix in X509_get_pubkey_parameters(3):schwarze2021-11-261-3/+5
| | | | | | | | | | If EVP_PKEY_copy_parameters(3) fails - among other reasons, this may happen when out of memory - the pkey argument and/or the chain argument will not contain all the desired parameters after returning. Consequently, report the failure to the caller rather than silently ignoring it. OK tb@
* Simplify the code in X509_get_pubkey_parameters(3)schwarze2021-11-261-8/+4
| | | | | | | | | | | | | | | by using X509_get0_pubkey(3) instead of X509_get_pubkey(3); no functional change. OK tb@ This is similar to the relevant part of the follwoing commit from the OpenSSL 1.1.1 branch, which is still under a free licence, but without the bug that commit introduced into this function in OpenSSL: commit c01ff880d47392b82cce2f93ac4a9bb8c68f8cc7 Author: Dr. Stephen Henson <steve@openssl.org> Date: Mon Dec 14 13:13:32 2015 +0000
* Add certificate transparency methods to the standard extensions.tb2021-11-241-1/+7
| | | | | | | | This way, CT extensions in certs will be parsed by the new CT code when they are encountered. This gets rid of a lot of gibberish when looking at a cert with 'openssl x509 -text -noout -in server.pem' ok beck jsing
* In some situations, the verifier would discard the error on an unvalidatedbeck2021-11-243-50/+91
| | | | | | certificte chain. This would happen when the verification callback was in use, instructing the verifier to continue unconditionally. This could lead to incorrect decisions being made in software.
* minor KNF improvement, changing only whitespace, no code change:schwarze2021-11-191-4/+4
| | | | | | say: return_type *function_name(args); not: return_type* function_name (args); OK tb@
* As long as X509_OBJECT_free_contents(3) is a public API function,schwarze2021-11-191-1/+3
| | | | | | | | | | | | | | | make sure it fully re-initializes the object rather than leaving behind a stale pointer and a stale type in the object. The old behaviour was dangerous because X509_OBJECT_get_type(3) would then return the stale type to the user and one of X509_OBJECT_get0_X509(3) or X509_OBJECT_get0_X509_CRL(3) would then return the stale pointer to the user, provoking a use-after-free bug in the application program. Having these functions return X509_LU_NONE and NULL is better because those are the documented return values for these functions when the object is empty. OK tb@
* Put curly brace on the correct line.jsing2021-11-141-2/+3
|
* Fix a bug in check_crl_time() that could result in incompleteschwarze2021-11-131-8/+8
| | | | | | | | | | | | | | | | | | | | | | | verification, accepting CRLs that ought to be rejected, if an unusual combination of verification flags was specified. If time verification was explicitly requested with X509_V_FLAG_USE_CHECK_TIME, it was skipped on CRLs if X509_V_FLAG_NO_CHECK_TIME was also set, even though the former is documented to override the latter both in the OpenSSL and in the LibreSSL X509_VERIFY_PARAM_set_flags(3) manual page. The same bug in x509_check_cert_time() was already fixed by beck@ in rev. 1.57 on 2017/01/20. This syncs the beginning of the function check_crl_time() with the OpenSSL 1.1.1 branch, which is still under a free license. OK beck@ This teaches that having too many flags and options is bad because they breed bugs, and even more so if they are poorly designed to override each other in surprising ways.
* Merge a few additional X509error(ERR_R_MALLOC_FAILURE) callsschwarze2021-11-101-39/+28
| | | | | | | | | | | | | | | | | | and various style improvements from the OpenSSL 1.1.1 branch, which is still under a free license. - No need to #include <openssl/lhash.h>. - BUF_MEM_free(3) and sk_pop_free(3) can handle NULL. - sk_value(3) can handle -1. - Test pointers with "== NULL" rather than with "!". - Use the safer "p = malloc(sizeof(*p))" idiom. - return is not a function. - Delete very wrong commented out code. Including parts of the these commits from the 2015 to 2018 time range: 25aaa98a b4faea50 90945fa3 f32b0abe 26a7d938 7fcdbd83 208056b2 5b37fef0 Requested by and OK tb@.
* If X509_load_cert_crl_file(3) does not find any certificatesschwarze2021-11-103-3/+7
| | | | | | | | | | | | | | | | | | and/or CRLs in the PEM input file (for example, if the file is empty), provide an error message in addition to returning 0. This merges another part of this OpenSSL commit, which is still under a free license: commit c0452248ea1a59a41023a4765ef7d9825e80a62b Author: Rich Salz <rsalz@openssl.org> Date: Thu Apr 20 15:33:42 2017 -0400 I did *not* add the similar message types X509_R_NO_CERTIFICATE_FOUND and X509_R_NO_CRL_FOUND because both code inspection and testing have shown that the code generating them is unreachable. OK tb@
* Sync some code style improvements from the OpenSSL 1.1.1 branch,schwarze2021-11-101-10/+9
| | | | | | | | | | | | | | which is still under a free license. No functional change. - No need to #include <openssl/lhash.h> here. - return is not a function. - Do not use the pointless macro BIO_s_file_internal(). - No need to check for NULL before X509_CRL_free(3). This includes parts of the following OpenSSL commits from the 2015 to 2017 timeframe: 222561fe, 9982cbbb, f32b0abe, 26a7d938 OK tb@
* Merge two bug fixes from the OpenSSL 1.1.1 branch, which is stillschwarze2021-11-101-8/+4
| | | | | | | | | | | | | | | | | | | | | | | | | under a free license: 1. If the three X509_load_*(3) functions are called with a NULL file argument, do not return 1 to the caller because the return value 1 means "i loaded one certificate or CRL into the store". 2. When calling PEM load functions, do not ask the user for a password in an interactive manner. This includes parts of the following commits: commit c0452248ea1a59a41023a4765ef7d9825e80a62b Author: Rich Salz <rsalz@openssl.org> Date: Thu Apr 20 15:33:42 2017 -0400 Message: [...] Remove NULL checks and allow a segv to occur. [...] commit db854bb14a7010712cfc02861731399b1b587474 Author: Bernd Edlinger <bernd.edlinger@hotmail.de> Date: Mon Aug 7 18:02:53 2017 +0200 Message: Avoid surpising password dialog in X509 file lookup. OK tb@
* In X509_STORE_CTX, rename the X509_STORE store rather than ctx.tb2021-11-073-15/+15
| | | | ok gnezdo jsing
* In X509_STORE_CTX rename the misnamed last_untrusted to num_untrustedtb2021-11-073-17/+17
| | | | ok jsing
* Start cleaning up X509_STORE_get1_issuer()tb2021-11-061-15/+37
| | | | | | | | | | | | | | Get rid of the last X509_OBJECT_free_contents() call by moving the object from the stack to the heap. I deliberately kept the obj variable to keep obj and pobj separate. Rename the out parameter from issuer to out_issuer to ensure that we only assign it when we have acquired a reference that we can return. Add a new X509 *issuer. In the first part of the function, acquire an extra reference before check_issuer/check_time. In the second part of the function, acquire a reference inside the lock to avoid a race. Deal with ret only in one place. ok jsing
* In X509_STORE_get1_issuer() do not call the verify callback fromtb2021-11-061-3/+3
| | | | | | x509_check_cert_time(). Matches a change made in OpenSSL 70dd3c65. ok jsing
* Refactor X509_STORE_get1_certs()tb2021-11-061-27/+30
| | | | | | | | Split the retrieval of the certs in the store's cache that match the desired subject into a separate function. This greatly simplifies locking, error handling and the flow of the function. with/ok jsing
* First pass of streamlining X509_STORE_get1_{certs,crls}()tb2021-11-051-66/+77
| | | | | | | | | | | | | | These functions are quite messy. On top of the tricky logic querying the cache, then refreshing the cache (unconditionally or not), then querying again, then extracting a list of certs/crls and bumping their refcounts, things are intermixed with locking and needlessly early allocations that then need to be cleaned up again. Use X509_STORE_CTX_get_obj_by_subject() to avoid using an object on the stack and defer allocation of the returned stack of certs to later. Flatten the logic a bit and prepare for further refactoring. ok jsing
* Trade an abort() neutered by a comment for a blank line elsewhere.tb2021-11-051-2/+2
|
* Clean up X509_STORE_add_{cert,crl}().tb2021-11-051-69/+41
| | | | | | | | | | | | | | | | | | | | Add a X509_STORE_add_object() function that adds an X509 object to the store and takes care of locking and cleaning up. This way we can set up an X509_OBJECT for both the cert and CRL case and hand over to the new function. There is one intentional change of behavior: if there is an attempt to add an object which is already present in the store, succeed instead of throwing an error. This makes sense and is also the OpenSSL behavior. As pointed out by jsing, this is a partial fix for the long standing GH issue #100 on libtls where connections would fail if the store contains duplicate certificates. Also: remove the internal X509_OBJECT_dec_ref_count(), which is no longer used. ok jsing
* Unify variable names in X509_STORE_{free,up_ref,add_lookup}().tb2021-11-051-29/+26
| | | | | | simplify the flow of X509_add_lookup(). ok jsing
generated by cgit v1.2.3-55-g6feb (git 2.39.1) at 2025-08-27 07:27:35 +0000