Fix issue with _pretty_size with large numbers of bytes

See merge request !799

Replaces the custom one which was erroneously added during development.
Fixes #746.

_sandboxremote.py: Use the standard SandboxError

Closes #746

See merge request !931

Update our instructions/examples regarding how we declare remote caches

See merge request !932

There are arguments for and against using onerror - see issue #735
for details.

Use _force_rmtree instead of custom code.

Closes #735

See merge request !927

Valentin David authored Nov 06, 2018 and richardmaw-codethink committed Nov 07, 2018

Root directory was marked as a non-artifact mount, so not using
SafeHardLink.  However integration commands executed with write access
to the root directory.

Fixes #749

Fix cache corruption by scripts when layout and integration commands are used

Closes #749

See merge request !934

Valentin David authored Nov 01, 2018 and Valentin David committed Nov 08, 2018

Fixes #645.

Fix infinite recursion in default strip debug command

Closes #645

See merge request !918

Benjamin Schubert authored Nov 07, 2018 and Benjamin Schubert committed Nov 08, 2018

Remove the bwraps checks from _site.py and put them in platform.linux
which is the only place where they are run.

This allows the removal of a double level of caching, making
reasoning about tests easier

Benjamin Schubert authored Nov 01, 2018 and Benjamin Schubert committed Nov 08, 2018

This removes the `_instance` on the platform object that we use
for caching and not recreating the object everytime at the start
of every test.

This is to ensure our tests share the least amount of state.

The performance penalty is from 5 to 10% accross the whole test suite.
The readings were done 5 times for each before and after the change
and on the same computer.

Benjamin Schubert authored Nov 01, 2018 and Benjamin Schubert committed Nov 08, 2018

This adds a `reason` to the SandboxEror thrown in sandboxdummy
to be able to understand where the error comes from

test behavior when there is no sandbox on Linux

Closes #696 and #736

See merge request !921

The issue was introduced by 6ccfab0b.

Fix bug with root mounted as non-artifact in script plugin.

Closes #756 and #749

See merge request !941

We were indiscriminately retrieving the node's provenance data in the
`node_get()` function which was accounting for approximately a third of
the total runtime of `node_get()` which dominates pre-scheduler time in
`bst build`.  This change ameliorates that situation by only retrieving
the provenance data when it's actually needed.

Signed-off-by: Daniel Silverstone <daniel.silverstone@codethink.co.uk>

We know that nodes are typically one of:
str, list, dict, bool, tuple, NoneType or our ChainMap
Of these, dict and ChainMap are Mapping, only list is list
and the rest are returned unchanged.  We can reduce/defer our use
of isinstance here, dramatically, improving performance.

Signed-off-by: Daniel Silverstone <daniel.silverstone@codethink.co.uk>

In order to support a case where the caller already has pre-sanitized input
this variant of the cache key generator needs to exist.

Signed-off-by: Daniel Silverstone <daniel.silverstone@codethink.co.uk>

In order to reduce the effort we spend in `_yaml.node_sanitize()` take
advantage of how `OrderedDict` works and pre-sanitize the majority of
the cache-key once.  This approximately halves the amount of effort we
spend in `_yaml.node_sanitize()` in pre-scheduler build scenarios.

Signed-off-by: Daniel Silverstone <daniel.silverstone@codethink.co.uk>