Call CRYPTO_set_locking_callback #352

public · 2013-12-26T17:14:29Z

So that we actually get some thread safety.

The cross platform locking stuff is based on the PyPy code and is entirely untested on Windows. This does fix the crash in my test case described in #330 on Linux.

public · 2013-12-26T17:14:47Z

Do we want a unit test that spawn a few threads and tries to cause bad things to happen? It usually crashes pretty fast on Linux if things aren't working but @reaperhulk said it didn't crash on Windows anyway :-/

coveralls · 2013-12-26T17:18:26Z

Coverage remained the same when pulling 45e673a on public:thread-safe into b645521 on pyca:master.

jenkins-cryptography · 2013-12-26T17:18:47Z

Can one of the admins verify this patch?

jenkins-cryptography · 2013-12-26T17:18:50Z

Can one of the admins verify this patch?

reaperhulk · 2013-12-26T17:19:08Z

add to whitelist

reaperhulk · 2013-12-26T17:23:04Z

I'd definitely like to see a test, but even with #330 it usually takes 20-30 seconds to crash on OS X. We could create a mark for slow tests if we wanted to start splitting this sort of thing out from the primary test suite? Would need to create a jenkins job to run the slow tests exclusively then...

jenkins-cryptography · 2013-12-26T17:26:18Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/68/

jenkins-cryptography · 2013-12-26T17:36:14Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/69/

coveralls · 2013-12-26T17:39:09Z

Coverage remained the same when pulling 057bff4 on public:thread-safe into b645521 on pyca:master.

coveralls · 2013-12-26T17:59:54Z

Coverage remained the same when pulling 057bff4 on public:thread-safe into b645521 on pyca:master.

coveralls · 2013-12-26T18:20:11Z

Coverage remained the same when pulling 057bff4 on public:thread-safe into b645521 on pyca:master.

jenkins-cryptography · 2013-12-26T18:38:02Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/70/

coveralls · 2013-12-26T18:39:53Z

Coverage remained the same when pulling b1f38cf on public:thread-safe into b645521 on pyca:master.

coveralls · 2013-12-26T19:01:10Z

Coverage remained the same when pulling b1f38cf on public:thread-safe into b645521 on pyca:master.

public · 2013-12-26T19:13:00Z

I think I can safely get rid of the sem_init version and just use the pthread_mutex on Linux now but I'm not really sure why CPython and PyPy both have implementations for each of them.

coveralls · 2013-12-26T19:58:14Z

Coverage remained the same when pulling 5e6949d on public:thread-safe into b645521 on pyca:master.

jenkins-cryptography · 2013-12-26T20:05:17Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/71/

jenkins-cryptography · 2013-12-26T20:05:52Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/72/

coveralls · 2013-12-26T20:12:25Z

Coverage remained the same when pulling afe1007 on public:thread-safe into b645521 on pyca:master.

coveralls · 2013-12-30T23:15:36Z

Coverage decreased (-0.11%) when pulling 5a178ef on public:thread-safe into b645521 on pyca:master.

public · 2013-12-30T23:18:59Z

I've added attempting to import _ssl to get Python to set it's handler. This also conditionally sets our handler if Python didn't for some reason. I'm not sure this buys us anything but making things more complex. _ssl won't reinit it's lock handler ever after the first import so simply immediately overwriting it's handler is quite safe.

I guess we save CRYPTO_num_locks() * sizeof(lock) memory though :)

jenkins-cryptography · 2013-12-30T23:22:33Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/118/

jenkins-cryptography · 2013-12-30T23:37:41Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/119/

coveralls · 2013-12-30T23:38:26Z

Coverage remained the same when pulling 38a853e on public:thread-safe into fbd7ffc on pyca:master.

jenkins-cryptography · 2013-12-30T23:53:05Z

Test FAILed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/120/

coveralls · 2013-12-30T23:59:21Z

Coverage decreased (-0.11%) when pulling 6a3c4de on public:thread-safe into fbd7ffc on pyca:master.

jenkins-cryptography · 2013-12-31T00:00:49Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/121/

coveralls · 2013-12-31T00:20:25Z

Coverage decreased (-0.11%) when pulling 6a3c4de on public:thread-safe into fbd7ffc on pyca:master.

coveralls · 2013-12-31T00:41:19Z

Coverage remained the same when pulling 80108a4 on public:thread-safe into fbd7ffc on pyca:master.

jenkins-cryptography · 2013-12-31T00:42:52Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/122/

alex · 2014-01-03T04:18:41Z

FYI this need to be rebased on master.

coveralls · 2014-01-03T14:09:33Z

Coverage remained the same when pulling de6f553 on public:thread-safe into adbea0d on pyca:master.

jenkins-cryptography · 2014-01-03T14:19:49Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/172/

jenkins-cryptography · 2014-01-03T16:39:02Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/180/

coveralls · 2014-01-03T18:00:23Z

Coverage remained the same when pulling 12ffba3 on public:thread-safe into adbea0d on pyca:master.

public · 2014-01-07T14:32:14Z

Did you get anywhere with using this in pyOpenSSL @exarkun ?

hynek · 2014-01-18T10:42:05Z

Given pyca/pyopenssl#7 I believe this should be high-priority.

So that we actually get some thread safety

because their sem_init is implemented as exit(1)

pthread_mutex works on Linux anyway and isn't obviously any better at coping with fork() issues.

Probably not particularly great at finding issues on all platforms. Notably Windows doesn't seem to crash without it anyway.

Adds quite a bit of complexity, might be better to set ours unconditionally but still import _ssl so that it doesn't overwrite ours at some point later.

Makes it a bit easier to read.

jenkins-cryptography · 2014-01-19T11:07:22Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/373/

jenkins-cryptography · 2014-01-19T11:18:09Z

Test PASSed.
Refer to this link for build results: http://jenkins.cryptography.io/job/cryptography-pr-experimental/374/

reaperhulk · 2014-01-20T15:47:53Z

Latest build returns these errors when attempting to compile on windows: http://bpaste.net/show/prfcOee90CjIL85YbuCz/

public · 2014-01-20T16:33:15Z

Weird given that it worked before. I think I know what's wrong though, will sort it out once I get a Windows VM to play with :)

alex · 2014-01-24T15:42:42Z

Closing in favor of #492

complete the flake8-ification of this file

public mentioned this pull request Jan 1, 2014

expose num_locks and {get,set}_{id,locking}_callback #387

Merged

public added 9 commits January 19, 2014 10:47

Call CRYPTO_set_locking_callback

b121816

So that we actually get some thread safety

Use the same platform ifdef check as PyPy

2dff72a

Add pthread_mutex mode for __APPLE__ Computers

1ae4e06

because their sem_init is implemented as exit(1)

Remove the sem_init implementation

d1d289b

pthread_mutex works on Linux anyway and isn't obviously any better at coping with fork() issues.

Test case for the locking stuff.

dfb3fd6

Probably not particularly great at finding issues on all platforms. Notably Windows doesn't seem to crash without it anyway.

Only init our locking function if Python didnt

de2c74b

Adds quite a bit of complexity, might be better to set ours unconditionally but still import _ssl so that it doesn't overwrite ours at some point later.

Update threading stuff for split backend/bindings

39ef7f6

Fix some compiler warnings

fff3675

Split the windows and posix lock backends

0b47203

Makes it a bit easier to read.

public mentioned this pull request Jan 21, 2014

Python implementation of OpenSSL locking callback #492

Merged

alex closed this Jan 24, 2014

joerichter-stash pushed a commit to kiwigrid/cryptography that referenced this pull request Nov 15, 2017

Merge pull request pyca#352 from alex/finish-test-crypto

2cfa855

complete the flake8-ification of this file

github-actions bot locked as resolved and limited conversation to collaborators Oct 10, 2020

Call CRYPTO_set_locking_callback #352

Call CRYPTO_set_locking_callback #352

Conversation

public commented Dec 26, 2013

public commented Dec 26, 2013

coveralls commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

reaperhulk commented Dec 26, 2013

reaperhulk commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

coveralls commented Dec 26, 2013

coveralls commented Dec 26, 2013

coveralls commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

coveralls commented Dec 26, 2013

coveralls commented Dec 26, 2013

public commented Dec 26, 2013

coveralls commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

jenkins-cryptography commented Dec 26, 2013

coveralls commented Dec 26, 2013

coveralls commented Dec 30, 2013

public commented Dec 30, 2013

jenkins-cryptography commented Dec 30, 2013

jenkins-cryptography commented Dec 30, 2013

coveralls commented Dec 30, 2013

jenkins-cryptography commented Dec 30, 2013

coveralls commented Dec 30, 2013

jenkins-cryptography commented Dec 31, 2013

coveralls commented Dec 31, 2013

coveralls commented Dec 31, 2013

jenkins-cryptography commented Dec 31, 2013

alex commented Jan 3, 2014

coveralls commented Jan 3, 2014

jenkins-cryptography commented Jan 3, 2014

jenkins-cryptography commented Jan 3, 2014

coveralls commented Jan 3, 2014

public commented Jan 7, 2014

hynek commented Jan 18, 2014

jenkins-cryptography commented Jan 19, 2014

jenkins-cryptography commented Jan 19, 2014

reaperhulk commented Jan 20, 2014

public commented Jan 20, 2014

alex commented Jan 24, 2014