[wip] combine kernel and kernel module cataloging

Signed-off-by: Alex Goodman <alex.goodman@anchore.com>
Signed-off-by: Avi Deitcher <avi@deitcher.net>

syft/pkg/cataloger/kernel/parse_kernel_module_file.go

@@ -3,21 +3,27 @@ package kernel
 import (
 	"debug/elf"
 	"fmt"
-	"github.com/anchore/syft/internal/log"
+	"regexp"
 	"strings"
 
+	"github.com/anchore/syft/internal/log"
 	"github.com/anchore/syft/syft/pkg"
 	"github.com/anchore/syft/syft/pkg/cataloger/internal/unionreader"
 	"github.com/anchore/syft/syft/source"
 )
 
+var (
+	kernelModulePathRE = regexp.MustCompile(`^(.*lib/modules/([^/]+)/)`)
+)
+
 func findKernelModules(resolver source.FileResolver, pkgLocations *source.LocationSet) ([]pkg.KernelModuleMetadata, error) {
 	locations, err := resolver.FilesByGlob("**/*.ko")
 	if err != nil {
 		return nil, err
 	}
 
 	// note: all collections in the metadata must be allocated
+	foundPaths := make(map[string]bool)
 	ret := make([]pkg.KernelModuleMetadata, 0)
 	for _, location := range locations {
 		contentReader, err := resolver.FileContentsByLocation(location)
@@ -35,9 +41,18 @@ func findKernelModules(resolver source.FileResolver, pkgLocations *source.Locati
 		if metadata == nil {
 			continue
 		}
-		pkgLocations.Add(location)
+		// get the base path
+		parts := kernelModulePathRE.FindStringSubmatch(location.RealPath)
+		if len(parts) > 1 {
+			basePath := parts[1]
+			if _, ok := foundPaths[basePath]; !ok {
+				foundPaths[basePath] = true
+				pkgLocations.Add(source.NewLocation(basePath))
+			}
+		}
 		ret = append(ret, *metadata)
 	}
+	// kernel modules have a common root, usually /lib/modules/<version>/kernel, so find the common base path
 	return ret, nil
 }
 

syft/pkg/kernel_package_metadata.go

@@ -1,5 +1,11 @@
 package pkg
 
+import (
+	"sort"
+
+	"github.com/scylladb/go-set/strset"
+)
+
 // KernelPackageMetadata represents all captured data for a Linux kernel
 type KernelPackageMetadata struct {
 	Name            string                 `mapstructure:"name" json:"name"`
@@ -16,6 +22,18 @@ type KernelPackageMetadata struct {
 	Modules         []KernelModuleMetadata `mapstructure:"modules" json:"modules"`
 }
 
+func (m KernelPackageMetadata) OwnedFiles() (result []string) {
+	s := strset.New()
+	for _, m := range m.Modules {
+		if m.Path != "" {
+			s.Add(m.Path)
+		}
+	}
+	result = s.List()
+	sort.Strings(result)
+	return
+}
+
 type KernelModuleMetadata struct {
 	KernelVersion string                           `mapstructure:"kernelVersion" json:"kernelVersion,omitempty"`
 	VersionMagic  string                           `mapstructure:"versionMagic" json:"versionMagic,omitempty"`