cache: create and use a RecipeInfo class

This class holds the particular pieces of information about a recipe which are
needed for runqueue to do its job.

By using it, I think we improve code clarity, reduce method sizes, reduce
overuse of primitive types, and prepare for parallel parsing.  In addition,
this ditches the leaky abstraction whereby bb.cache attempted to hide the
difference between cached data and a full recipe parse.  This was a remnant
from the way things used to be done, and the code using it had to know the
difference anyway.  If we choose to reimplement caching of the full recipes,
we can do it in bb.parse, in a completely transparent way.

(Bitbake rev: 992cc252452221f5f23575e50eb67528b2838fdb)

Signed-off-by: Chris Larson <chris_larson@mentor.com>
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>

2 files changed, 197 insertions, 175 deletions

diff --git a/bitbake/lib/bb/cache.py b/bitbake/lib/bb/cache.py
index 707d81c97..3e2f698cf 100644
--- a/bitbake/lib/bb/cache.py
+++ b/bitbake/lib/bb/cache.py
@@ -30,7 +30,7 @@
 
 import os
 import logging
-from collections import defaultdict
+from collections import defaultdict, namedtuple
 import bb.data
 import bb.utils
 
@@ -43,7 +43,104 @@ except ImportError:
     logger.info("Importing cPickle failed. "
                 "Falling back to a very slow implementation.")
 
-__cache_version__ = "132"
+__cache_version__ = "133"
+
+recipe_fields = (
+    'pn',
+    'pv',
+    'pr',
+    'pe',
+    'defaultpref',
+    'depends',
+    'provides',
+    'task_deps',
+    'stamp',
+    'broken',
+    'not_world',
+    'skipped',
+    'timestamp',
+    'packages',
+    'packages_dynamic',
+    'rdepends',
+    'rdepends_pkg',
+    'rprovides',
+    'rprovides_pkg',
+    'rrecommends',
+    'rrecommends_pkg',
+    'nocache',
+    'variants',
+    'file_depends',
+    'tasks',
+    'basetaskhashes',
+    'hashfilename',
+)
+
+
+class RecipeInfo(namedtuple('RecipeInfo', recipe_fields)):
+    __slots__ = ()
+
+    @classmethod
+    def listvar(cls, var, metadata):
+        return cls.getvar(var, metadata).split()
+
+    @classmethod
+    def intvar(cls, var, metadata):
+        return int(cls.getvar(var, metadata) or 0)
+
+    @classmethod
+    def depvar(cls, var, metadata):
+        return bb.utils.explode_deps(cls.getvar(var, metadata))
+
+    @classmethod
+    def pkgvar(cls, var, packages, metadata):
+        return dict((pkg, cls.depvar("%s_%s" % (var, pkg), metadata))
+                    for pkg in packages)
+
+    @classmethod
+    def taskvar(cls, var, tasks, metadata):
+        return dict((task, cls.getvar("%s_task-%s" % (var, task), metadata))
+                    for task in tasks)
+    @classmethod
+    def getvar(cls, var, metadata):
+        return metadata.getVar(var, True) or ''
+
+    @classmethod
+    def from_metadata(cls, filename, metadata):
+        tasks = metadata.getVar('__BBTASKS', False)
+
+        packages = cls.listvar('PACKAGES', metadata)
+        return RecipeInfo(
+            tasks            = tasks,
+            basetaskhashes   = cls.taskvar('BB_BASEHASH', tasks, metadata),
+            hashfilename     = cls.getvar('BB_HASHFILENAME', metadata),
+
+            file_depends     = metadata.getVar('__depends', False),
+            task_deps        = metadata.getVar('_task_deps', False) or
+                               {'tasks': [], 'parents': {}},
+            variants         = cls.listvar('__VARIANTS', metadata) + [''],
+
+            skipped          = cls.getvar('__SKIPPED', metadata),
+            timestamp        = bb.parse.cached_mtime(filename),
+            packages         = packages,
+            pn               = cls.getvar('PN', metadata),
+            pe               = cls.getvar('PE', metadata),
+            pv               = cls.getvar('PV', metadata),
+            pr               = cls.getvar('PR', metadata),
+            nocache          = cls.getvar('__BB_DONT_CACHE', metadata),
+            defaultpref      = cls.intvar('DEFAULT_PREFERENCE', metadata),
+            broken           = cls.getvar('BROKEN', metadata),
+            not_world        = cls.getvar('EXCLUDE_FROM_WORLD', metadata),
+            stamp            = cls.getvar('STAMP', metadata),
+            packages_dynamic = cls.listvar('PACKAGES_DYNAMIC', metadata),
+            depends          = cls.depvar('DEPENDS', metadata),
+            provides         = cls.depvar('PROVIDES', metadata),
+            rdepends         = cls.depvar('RDEPENDS', metadata),
+            rprovides        = cls.depvar('RPROVIDES', metadata),
+            rrecommends      = cls.depvar('RRECOMMENDS', metadata),
+            rprovides_pkg    = cls.pkgvar('RPROVIDES', packages, metadata),
+            rdepends_pkg     = cls.pkgvar('RDEPENDS', packages, metadata),
+            rrecommends_pkg  = cls.pkgvar('RRECOMMENDS', packages, metadata),
+        )
 
 
 class Cache(object):
@@ -99,56 +196,6 @@ class Cache(object):
             if os.path.isfile(self.cachefile):
                 logger.info("Out of date cache found, rebuilding...")
 
-    def getVar(self, var, fn, exp=0):
-        """
-        Gets the value of a variable
-        (similar to getVar in the data class)
-
-        There are two scenarios:
-          1. We have cached data - serve from depends_cache[fn]
-          2. We're learning what data to cache - serve from data
-             backend but add a copy of the data to the cache.
-        """
-        if fn in self.clean:
-            return self.depends_cache[fn].get(var)
-
-        self.depends_cache.setdefault(fn, {})
-
-        if fn != self.data_fn:
-            # We're trying to access data in the cache which doesn't exist
-            # yet setData hasn't been called to setup the right access
-            logger.error("data_fn %s and fn %s don't match", self.data_fn, fn)
-
-        self.cacheclean = False
-        result = bb.data.getVar(var, self.data, exp)
-        if result is not None:
-            self.depends_cache[fn][var] = result
-        return result
-
-    def setData(self, virtualfn, fn, data):
-        """
-        Called to prime bb_cache ready to learn which variables to cache.
-        Will be followed by calls to self.getVar which aren't cached
-        but can be fulfilled from self.data.
-        """
-        self.data_fn = virtualfn
-        self.data = data
-
-        # Make sure __depends makes the depends_cache
-        # If we're a virtual class we need to make sure all our depends are
-        # appended to the depends of fn.
-        depends = self.getVar("__depends", virtualfn) or set()
-        self.depends_cache.setdefault(fn, {})
-        if "__depends" not in self.depends_cache[fn] or not self.depends_cache[fn]["__depends"]:
-            self.depends_cache[fn]["__depends"] = depends
-        else:
-            self.depends_cache[fn]["__depends"].update(depends)
-
-        # Make sure the variants always make it into the cache too
-        self.getVar('__VARIANTS', virtualfn, True)
-
-        self.depends_cache[virtualfn]["CACHETIMESTAMP"] = bb.parse.cached_mtime(fn)
-
     @staticmethod
     def virtualfn2realfn(virtualfn):
         """
@@ -193,38 +240,39 @@ class Cache(object):
         to record the variables accessed.
         Return the cache status and whether the file was skipped when parsed
         """
-        skipped = 0
-        virtuals = 0
+        skipped, virtuals = 0, 0
 
         if fn not in self.checked:
             self.cacheValidUpdate(fn)
 
-        if self.cacheValid(fn):
-            multi = self.getVar('__VARIANTS', fn, True)
-            for cls in (multi or "").split() + [""]:
-                virtualfn = self.realfn2virtual(fn, cls)
-                if self.depends_cache[virtualfn].get("__SKIPPED"):
-                    skipped += 1
-                    logger.debug(1, "Skipping %s", virtualfn)
-                    continue
-                self.handle_data(virtualfn, cacheData)
-                virtuals += 1
-            return True, skipped, virtuals
-
-        logger.debug(1, "Parsing %s", fn)
-
-        bb_data = self.load_bbfile(fn, appends, cfgData)
-
-        for data in bb_data:
-            virtualfn = self.realfn2virtual(fn, data)
-            self.setData(virtualfn, fn, bb_data[data])
-            if self.getVar("__SKIPPED", virtualfn):
-                skipped += 1
+        cached = self.cacheValid(fn)
+        if not cached:
+            self.cacheclean = False
+            logger.debug(1, "Parsing %s", fn)
+            datastores = self.load_bbfile(fn, appends, cfgData)
+            depends = set()
+            for variant, data in sorted(datastores.iteritems(),
+                                        key=lambda i: i[0],
+                                        reverse=True):
+                virtualfn = self.realfn2virtual(fn, variant)
+                depends |= (data.getVar("__depends", False) or set())
+                if depends and not variant:
+                    data.setVar("__depends", depends)
+                info = RecipeInfo.from_metadata(fn, data)
+                self.depends_cache[virtualfn] = info
+
+        info = self.depends_cache[fn]
+        for variant in info.variants:
+            virtualfn = self.realfn2virtual(fn, variant)
+            vinfo = self.depends_cache[virtualfn]
+            if vinfo.skipped:
                 logger.debug(1, "Skipping %s", virtualfn)
+                skipped += 1
             else:
-                self.handle_data(virtualfn, cacheData)
+                cacheData.add_from_recipeinfo(virtualfn, vinfo)
                 virtuals += 1
-        return False, skipped, virtuals
+
+        return cached, skipped, virtuals
 
     def cacheValid(self, fn):
         """
@@ -266,14 +314,15 @@ class Cache(object):
             self.remove(fn)
             return False
 
+        info = self.depends_cache[fn]
         # Check the file's timestamp
-        if mtime != self.getVar("CACHETIMESTAMP", fn, True):
+        if mtime != info.timestamp:
             logger.debug(2, "Cache: %s changed", fn)
             self.remove(fn)
             return False
 
         # Check dependencies are still valid
-        depends = self.getVar("__depends", fn, True)
+        depends = info.file_depends
         if depends:
             for f, old_mtime in depends:
                 fmtime = bb.parse.cached_mtime_noerror(f)
@@ -290,20 +339,17 @@ class Cache(object):
                     self.remove(fn)
                     return False
 
-        self.clean.add(fn)
         invalid = False
-        # Mark extended class data as clean too
-        multi = self.getVar('__VARIANTS', fn, True)
-        for cls in (multi or "").split():
+        for cls in info.variants:
             virtualfn = self.realfn2virtual(fn, cls)
             self.clean.add(virtualfn)
-            if not virtualfn in self.depends_cache:
+            if virtualfn not in self.depends_cache:
                 logger.debug(2, "Cache: %s is not cached", virtualfn)
                 invalid = True
 
         # If any one of the variants is not present, mark as invalid for all
         if invalid:
-            for cls in (multi or "").split():
+            for cls in info.variants:
                 virtualfn = self.realfn2virtual(fn, cls)
                 if virtualfn in self.clean:
                     logger.debug(2, "Cache: Removing %s from cache", virtualfn)
@@ -332,7 +378,6 @@ class Cache(object):
         Save the cache
         Called from the parser when complete (or exiting)
         """
-        import copy
 
         if not self.has_cache:
             return
@@ -345,13 +390,12 @@ class Cache(object):
         version_data['CACHE_VER'] = __cache_version__
         version_data['BITBAKE_VER'] = bb.__version__
 
-        cache_data = copy.copy(self.depends_cache)
-        for fn in self.depends_cache:
-            if '__BB_DONT_CACHE' in self.depends_cache[fn] and self.depends_cache[fn]['__BB_DONT_CACHE']:
+        cache_data = dict(self.depends_cache)
+        for fn, info in self.depends_cache.iteritems():
+            if info.nocache:
                 logger.debug(2, "Not caching %s, marked as not cacheable", fn)
                 del cache_data[fn]
-            elif ('PV' in self.depends_cache[fn] and
-                  'SRCREVINACTION' in self.depends_cache[fn]['PV']):
+            elif info.pv and 'SRCREVINACTION' in info.pv:
                 logger.error("Not caching %s as it had SRCREVINACTION in PV. "
                              "Please report this bug", fn)
                 del cache_data[fn]
@@ -364,90 +408,14 @@ class Cache(object):
     def mtime(cachefile):
         return bb.parse.cached_mtime_noerror(cachefile)
 
-    def handle_data(self, file_name, cacheData):
+    def add(self, file_name, data, cacheData):
         """
         Save data we need into the cache
         """
-
-        pn = self.getVar('PN', file_name, True)
-        pe = self.getVar('PE', file_name, True) or "0"
-        pv = self.getVar('PV', file_name, True)
-        if 'SRCREVINACTION' in pv:
-            logger.info("Found SRCREVINACTION in PV (%s) or %s. Please report this bug.", pv, file_name)
-        pr = self.getVar('PR', file_name, True)
-        dp = int(self.getVar('DEFAULT_PREFERENCE', file_name, True) or "0")
-        depends = bb.utils.explode_deps(self.getVar("DEPENDS", file_name, True) or "")
-        packages = (self.getVar('PACKAGES', file_name, True) or "").split()
-        packages_dynamic = (self.getVar('PACKAGES_DYNAMIC', file_name, True) or "").split()
-        rprovides = (self.getVar("RPROVIDES", file_name, True) or "").split()
-
-        cacheData.task_deps[file_name] = self.getVar("_task_deps", file_name)
-
-        # build PackageName to FileName lookup table
-        cacheData.pkg_pn[pn].append(file_name)
-
-        cacheData.stamp[file_name] = self.getVar('STAMP', file_name, True)
-
-        cacheData.tasks[file_name] = self.getVar('__BBTASKS', file_name, True)
-        for t in cacheData.tasks[file_name]:
-            cacheData.basetaskhash[file_name + "." + t] = self.getVar("BB_BASEHASH_task-%s" % t, file_name, True)
-
-        # build FileName to PackageName lookup table
-        cacheData.pkg_fn[file_name] = pn
-        cacheData.pkg_pepvpr[file_name] = (pe, pv, pr)
-        cacheData.pkg_dp[file_name] = dp
-
-        provides = [pn]
-        for provide in (self.getVar("PROVIDES", file_name, True) or "").split():
-            if provide not in provides:
-                provides.append(provide)
-
-        # Build forward and reverse provider hashes
-        # Forward: virtual -> [filenames]
-        # Reverse: PN -> [virtuals]
-        cacheData.fn_provides[file_name] = provides
-        for provide in provides:
-            cacheData.providers[provide].append(file_name)
-            if not provide in cacheData.pn_provides[pn]:
-                cacheData.pn_provides[pn].append(provide)
-
-        for dep in depends:
-            if not dep in cacheData.deps[file_name]:
-                cacheData.deps[file_name].append(dep)
-            if not dep in cacheData.all_depends:
-                cacheData.all_depends.append(dep)
-
-        # Build reverse hash for PACKAGES, so runtime dependencies
-        # can be be resolved (RDEPENDS, RRECOMMENDS etc.)
-        for package in packages:
-            cacheData.packages[package].append(file_name)
-            rprovides += (self.getVar("RPROVIDES_%s" % package, file_name, 1) or "").split()
-
-        for package in packages_dynamic:
-            cacheData.packages_dynamic[package].append(file_name)
-
-        for rprovide in rprovides:
-            cacheData.rproviders[rprovide].append(file_name)
-
-        # Build hash of runtime depends and rececommends
-        rdepends = bb.utils.explode_deps(self.getVar('RDEPENDS', file_name, True) or "")
-        rrecommends = bb.utils.explode_deps(self.getVar('RRECOMMENDS', file_name, True) or "")
-        for package in packages + [pn]:
-            rdeps_pkg = bb.utils.explode_deps(self.getVar('RDEPENDS_%s' % package, file_name, True) or "")
-            cacheData.rundeps[file_name][package] = rdepends + rdeps_pkg
-            rrecs_pkg = bb.utils.explode_deps(self.getVar('RDEPENDS_%s' % package, file_name, True) or "")
-            cacheData.runrecs[file_name][package] = rrecommends + rrecs_pkg
-
-        # Collect files we may need for possible world-dep
-        # calculations
-        if not self.getVar('BROKEN', file_name, True) and not self.getVar('EXCLUDE_FROM_WORLD', file_name, True):
-            cacheData.possible_world.append(file_name)
-
-        cacheData.hashfn[file_name] = self.getVar('BB_HASHFILENAME', file_name, True)
-
-        # Touch this to make sure its in the cache
-        self.getVar('__BB_DONT_CACHE', file_name, True)
-        self.getVar('__VARIANTS', file_name, True)
+        realfn = self.virtualfn2realfn(file_name)[0]
+        info = RecipeInfo.from_metadata(realfn, data)
+        self.depends_cache[file_name] = info
+        cacheData.add_from_recipeinfo(file_name, info)
 
     @staticmethod
     def load_bbfile(bbfile, appends, config):
@@ -514,7 +482,6 @@ class CacheData(object):
     def __init__(self):
         """
         Direct cache variables
-        (from Cache.handle_data)
         """
         self.providers = defaultdict(list)
         self.rproviders = defaultdict(list)
@@ -547,3 +514,59 @@ class CacheData(object):
         self.world_target = set()
         self.bbfile_priority = {}
         self.bbfile_config_priorities = []
+
+    def add_from_recipeinfo(self, fn, info):
+        self.task_deps[fn] = info.task_deps
+        self.pkg_fn[fn] = info.pn
+        self.pkg_pn[info.pn].append(fn)
+        self.pkg_pepvpr[fn] = (info.pe, info.pv, info.pr)
+        self.pkg_dp[fn] = info.defaultpref
+        self.stamp[fn] = info.stamp
+
+        provides = [info.pn]
+        for provide in info.provides:
+            if provide not in provides:
+                provides.append(provide)
+        self.fn_provides[fn] = provides
+
+        for provide in provides:
+            self.providers[provide].append(fn)
+            if provide not in self.pn_provides[info.pn]:
+                self.pn_provides[info.pn].append(provide)
+
+        for dep in info.depends:
+            if dep not in self.deps[fn]:
+                self.deps[fn].append(dep)
+            if dep not in self.all_depends:
+                self.all_depends.append(dep)
+
+        rprovides = info.rprovides
+        for package in info.packages:
+            self.packages[package].append(fn)
+            rprovides += info.rprovides_pkg[package]
+
+        for package in info.packages_dynamic:
+            self.packages_dynamic[package].append(fn)
+
+        for rprovide in rprovides:
+            self.rproviders[rprovide].append(fn)
+
+        # Build hash of runtime depends and rececommends
+        for package in info.packages + [info.pn]:
+            rundeps, runrecs = list(info.rdepends), list(info.rrecommends)
+            if package in info.packages:
+                rundeps += info.rdepends_pkg[package]
+                runrecs += info.rrecommends_pkg[package]
+            self.rundeps[fn][package] = rundeps
+            self.runrecs[fn][package] = runrecs
+
+        # Collect files we may need for possible world-dep
+        # calculations
+        if not info.broken and not info.not_world:
+            self.possible_world.append(fn)
+
+        self.hashfn[fn] = info.hashfilename
+
+        for task, taskhash in info.basetaskhashes.iteritems():
+            identifier = '%s.%s' % (fn, task)
+            self.basetaskhash[identifier] = taskhash
diff --git a/bitbake/lib/bb/cooker.py b/bitbake/lib/bb/cooker.py
index 2e139558b..6194919e4 100644
--- a/bitbake/lib/bb/cooker.py
+++ b/bitbake/lib/bb/cooker.py
@@ -648,8 +648,7 @@ class BBCooker:
 
         # Load data into the cache for fn and parse the loaded cache data
         the_data = self.bb_cache.loadDataFull(fn, self.get_file_appends(fn), self.configuration.data)
-        self.bb_cache.setData(fn, buildfile, the_data)
-        self.bb_cache.handle_data(fn, self.status)
+        self.bb_cache.add(fn, the_data, self.status)
 
         # Tweak some variables
         item = self.bb_cache.getVar('PN', fn, True)