Fix error-handling for non-SSL_get_error functions.

Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none

Committed: https://crrev.com/f225b26378120e2c3dfd28f3e451ffb2d4e8c819
Cr-Commit-Position: refs/heads/master@{#419006}

[davidben, 2016-09-15 21:04:16]

The CQ bit was checked by davidben@chromium.org to run a CQ dry run

[davidben, 2016-09-15 21:04:30]

davidben@chromium.org changed reviewers:
+ svaldez@chromium.org

[commit-bot: I haz the power, 2016-09-15 21:04:53]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[davidben, 2016-09-15 21:05:00]

Description was changed from

==========
Simplify error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none
==========

to

==========
Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none
==========

[svaldez, 2016-09-15 21:06:12]

lgtm

[davidben, 2016-09-15 21:31:31]

The CQ bit was unchecked by davidben@chromium.org

[davidben, 2016-09-15 21:31:33]

The CQ bit was checked by davidben@chromium.org

[commit-bot: I haz the power, 2016-09-15 21:32:27]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2016-09-15 22:09:41]

Description was changed from

==========
Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none
==========

to

==========
Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none
==========

[commit-bot: I haz the power, 2016-09-15 22:09:43]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-09-15 22:14:28]

Description was changed from

==========
Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none
==========

to

==========
Fix error-handling for non-SSL_get_error functions.

Some functions (the ones that do I/O) use this crazy three-level error
system (rv, SSL_get_error, and the error queue). Others are more
straightforward and simply use the error queue.

Using SSL_get_error here could result in state from other operations
getting in the way of things.

Since these don't actually do I/O and really can only fail on internal
error, just map to ERR_FAILED rather than try to have so many
MapOpenSSLError variants.

Also treat SSL_export_key_material's return value as a boolean. In
BoringSSL, the calling convention is simpler and documented as such.

BUG=none

Committed: https://crrev.com/f225b26378120e2c3dfd28f3e451ffb2d4e8c819
Cr-Commit-Position: refs/heads/master@{#419006}
==========

[commit-bot: I haz the power, 2016-09-15 22:14:29]

Patchset 1 (id:??) landed as
https://crrev.com/f225b26378120e2c3dfd28f3e451ffb2d4e8c819
Cr-Commit-Position: refs/heads/master@{#419006}