December 2016 - tor-commits - lists.torproject.org

[tor/release-0.2.9] Refactor poison_dir allocation and free in test_single_onion_poisoning
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit c100c5c69b4fd3b464b2395263e77cc6e1051ef3 Author: teor <teor2345(a)gmail.com> Date: Sat Dec 3 06:25:46 2016 +1100 Refactor poison_dir allocation and free in test_single_onion_poisoning This pattern is much less error-prone when future changes are made. --- src/test/test_hs.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/src/test/test_hs.c b/src/test/test_hs.c index f4ba7f9..1039c64 100644 --- a/src/test/test_hs.c +++ b/src/test/test_hs.c @@ -541,6 +541,7 @@ test_single_onion_poisoning(void *arg) rend_service_t *service_2 = tor_malloc_zero(sizeof(rend_service_t)); char *dir2 = tor_strdup(get_fname_rnd("test_hs_dir2")); smartlist_t *services = smartlist_new(); + char *poison_path = NULL; /* No services, no service to verify, no problem! */ mock_options->HiddenServiceSingleHopMode = 0; @@ -706,9 +707,9 @@ test_single_onion_poisoning(void *arg) /* Now remove the poisoning from the first service, and we have the opposite * problem. */ - char *poison_path = rend_service_sos_poison_path(service_1); + poison_path = rend_service_sos_poison_path(service_1); + tt_assert(poison_path); ret = unlink(poison_path); - tor_free(poison_path); tt_assert(ret == 0); /* Unpoisoned service directories with previous keys are ok, as are empty @@ -765,6 +766,7 @@ test_single_onion_poisoning(void *arg) tt_assert(ret == 0); done: + tor_free(poison_path); tor_free(dir1); tor_free(dir2); /* The test harness deletes the directories at exit */

1 0

[tor/release-0.2.9] Merge branch 'maint-0.2.9' into release-0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit b4c648c6c0755e915fefb33431c2865382f11cf8 Merge: 4178708 42ec60e Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:13:09 2016 -0500 Merge branch 'maint-0.2.9' into release-0.2.9 changes/bug20864 | 4 ++++ src/test/test_hs.c | 13 ++++++++----- 2 files changed, 12 insertions(+), 5 deletions(-)

1 0

[tor/release-0.2.9] Merge remote-tracking branch 'teor/bug20864-029' into maint-0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit a23c581d13c164752effa9ccfa840192187f1388 Merge: 0d0c7a1 ebf243b Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:10:48 2016 -0500 Merge remote-tracking branch 'teor/bug20864-029' into maint-0.2.9 changes/bug20864 | 4 ++++ src/test/test_hs.c | 13 ++++++++----- 2 files changed, 12 insertions(+), 5 deletions(-)

1 0

[tor/release-0.2.9] Fix changes file for 20864: 20638 _did_ get into 0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 42ec60ecfbff39c454945bc52def0df196ec38b8 Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:12:10 2016 -0500 Fix changes file for 20864: 20638 _did_ get into 0.2.9 --- changes/bug20864 | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/changes/bug20864 b/changes/bug20864 index 079e609..7b8c70f 100644 --- a/changes/bug20864 +++ b/changes/bug20864 @@ -1,4 +1,4 @@ - o Minor bugfixes (hidden services): + o Minor bugfixes (unit tests, hidden services): - Remove a double-free in the single onion service unit test. Stop ignoring a return value. Make future changes less error-prone. - Fixes bug 20864; bugfix on 20638; not in any released version of tor. + Fixes bug 20864; bugfix on 0.2.9.6-rc.

1 0

[tor/master] Merge branch 'maint-0.2.8' into maint-0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 0d0c7a11649db161070ab96310ff301be6dd704c Merge: 907cd8a 2a36541 Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:05:37 2016 -0500 Merge branch 'maint-0.2.8' into maint-0.2.9

1 0

[tor/master] Always Use EVP_aes_*_ctr() with openssl 1.1
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 2a365413ebc7ab0055bd4c2fc11697666fc90f7b Author: Nick Mathewson <nickm(a)torproject.org> Date: Sun Nov 6 21:01:25 2016 -0500 Always Use EVP_aes_*_ctr() with openssl 1.1 (OpenSSL 1.1 makes EVP_CIPHER_CTX opaque, _and_ adds acceleration for counter mode on more architectures. So it won't work if we try the older approach, and it might help if we try the newer one.) Fixes bug 20588. --- changes/bug20588 | 3 +++ src/common/aes.c | 9 ++++++++- 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/changes/bug20588 b/changes/bug20588 new file mode 100644 index 0000000..be199b2 --- /dev/null +++ b/changes/bug20588 @@ -0,0 +1,3 @@ + o Minor bugfixes (portability): + - Fix compilation with OpenSSL 1.1 and less commonly-used + CPU architectures. Closes ticket 20588. diff --git a/src/common/aes.c b/src/common/aes.c index 15970a7..8edfc5d 100644 --- a/src/common/aes.c +++ b/src/common/aes.c @@ -73,7 +73,14 @@ * gives us, and the best possible counter-mode implementation, and combine * them. */ -#if OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,0,1) && \ +#if OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,1,0) + +/* With newer OpenSSL versions, the older fallback modes don't compile. So + * don't use them, even if we lack specific acceleration. */ + +#define USE_EVP_AES_CTR + +#elif OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,0,1) && \ (defined(__i386) || defined(__i386__) || defined(_M_IX86) || \ defined(__x86_64) || defined(__x86_64__) || \ defined(_M_AMD64) || defined(_M_X64) || defined(__INTEL__)) \

1 0

[tor/master] Merge branch 'maint-0.2.9'
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 6eb1b4bd9cbf16339b30c2531dd1997291cc0d4c Merge: 8a0ea3e 0d0c7a1 Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:06:36 2016 -0500 Merge branch 'maint-0.2.9'

1 0

[tor/maint-0.2.9] Merge branch 'maint-0.2.8' into maint-0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 0d0c7a11649db161070ab96310ff301be6dd704c Merge: 907cd8a 2a36541 Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:05:37 2016 -0500 Merge branch 'maint-0.2.8' into maint-0.2.9

1 0

[tor/release-0.2.8] Always Use EVP_aes_*_ctr() with openssl 1.1
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 2a365413ebc7ab0055bd4c2fc11697666fc90f7b Author: Nick Mathewson <nickm(a)torproject.org> Date: Sun Nov 6 21:01:25 2016 -0500 Always Use EVP_aes_*_ctr() with openssl 1.1 (OpenSSL 1.1 makes EVP_CIPHER_CTX opaque, _and_ adds acceleration for counter mode on more architectures. So it won't work if we try the older approach, and it might help if we try the newer one.) Fixes bug 20588. --- changes/bug20588 | 3 +++ src/common/aes.c | 9 ++++++++- 2 files changed, 11 insertions(+), 1 deletion(-) diff --git a/changes/bug20588 b/changes/bug20588 new file mode 100644 index 0000000..be199b2 --- /dev/null +++ b/changes/bug20588 @@ -0,0 +1,3 @@ + o Minor bugfixes (portability): + - Fix compilation with OpenSSL 1.1 and less commonly-used + CPU architectures. Closes ticket 20588. diff --git a/src/common/aes.c b/src/common/aes.c index 15970a7..8edfc5d 100644 --- a/src/common/aes.c +++ b/src/common/aes.c @@ -73,7 +73,14 @@ * gives us, and the best possible counter-mode implementation, and combine * them. */ -#if OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,0,1) && \ +#if OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,1,0) + +/* With newer OpenSSL versions, the older fallback modes don't compile. So + * don't use them, even if we lack specific acceleration. */ + +#define USE_EVP_AES_CTR + +#elif OPENSSL_VERSION_NUMBER >= OPENSSL_V_NOPATCH(1,0,1) && \ (defined(__i386) || defined(__i386__) || defined(_M_IX86) || \ defined(__x86_64) || defined(__x86_64__) || \ defined(_M_AMD64) || defined(_M_X64) || defined(__INTEL__)) \

1 0

[tor/release-0.2.9] Merge branch 'maint-0.2.8' into maint-0.2.9
by nickm＠torproject.org 05 Dec '16

05 Dec '16

commit 0d0c7a11649db161070ab96310ff301be6dd704c Merge: 907cd8a 2a36541 Author: Nick Mathewson <nickm(a)torproject.org> Date: Mon Dec 5 08:05:37 2016 -0500 Merge branch 'maint-0.2.8' into maint-0.2.9

1 0