Message93324
Uhm... I'm sorry but actually I'm not sure about this patch anymore.
Now that I look at ssl.py again I'm noticing that send() is trapped in a
"while True" loop as well and the patch doesn't cover it.
Not sure if that has been added recently or it was already there at the
time I submitted the report but it's another thing that need to be
fixed.
Moreover, I'm sure that removing the "while" loop is good for non-
blocking sockets but what about blocking ones?
Are SSL_ERROR_WANT_READ and SSL_ERROR_WANT_WRITE supposed to be raised
in a blocking environment?
If they aren't then the current patch just needs to take care of send()
method too, then it's fine. |
|
Date |
User |
Action |
Args |
2009-09-29 19:47:33 | giampaolo.rodola | set | recipients:
+ giampaolo.rodola, barry, josiahcarlson, exarkun, amaury.forgeotdarc, janssen, josiah.carlson, ddvoinikov, qwavel |
2009-09-29 19:47:33 | giampaolo.rodola | set | messageid: <1254253653.38.0.148741514276.issue3890@psf.upfronthosting.co.za> |
2009-09-29 19:47:31 | giampaolo.rodola | link | issue3890 messages |
2009-09-29 19:47:31 | giampaolo.rodola | create | |
|