* fix(redact): reset redaction store between command runs
cli.Command() may be invoked more than once in the same process (e.g.
when syft is embedded as a library). The clio initializer calls
internal/redact.Set on every command execution, but the redact store is
process-global and Set panics when a store already exists, so the second
invocation dies with "replace existing redaction store (probably
unintentional)".
Add redact.Reset() to clear the previous run's store and call it in the
initializer before Set. The double-Set guard in Set is kept: an
unexpected second Set within a single run still panics.
Add TestAppClioSetupConfigInitializerCanRunMultipleTimes which runs the
initializer twice; it panics with the exact reported message on the old
code and passes with the fix.
Fixes#2285
Signed-off-by: JasonMetal <935216773@qq.com>
* fix(redact): release the redact store at the end of each run
Clear the global redact store in the post-run hook instead of resetting it
before every `Set`. The double-`Set` panic keeps its meaning: a store is only
present while a run is in flight, so a second run starting mid-run still
panics rather than splitting secrets across two stores and leaking the ones
in the dropped store. Sequential runs in the same process work since the
previous run releases its store on the way out.
A run only releases the store it set, so it never clears a store owned by
someone else.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: JasonMetal <935216773@qq.com>
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Co-authored-by: Alex Goodman <wagoodman@users.noreply.github.com>
* ci: run the full unit suite on windows via make unit
Drops the curated `unit:windows` subset; the windows job now bootstraps like linux and runs `make unit`. Exceptions get added as failures show up.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* ci: skip fixture refresh for windows unit tests
binny has no windows release, so the bootstrap action can't install the tools the fixture refresh needs.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: skip linux container fixtures when docker can't run them
Windows CI runners only run windows containers, so image fixtures go through `testutils.SkipWithoutLinuxContainers`. No behavior change on linux/mac.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: close leaked file handles and posix squashfs paths on windows
- output writers are now closed on early returns in `scan`/`convert` (open handles block cleanup on windows)
- `newSBOMMultiWriter` used `path.Dir` on host paths, so output subdirectories were never created on windows
- squashfs walk paths were built with `filepath.Join`
- skip a few posix-only test cases (dir perms, xdg fallback, go-rpmdb ndb handle leak upstream)
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(golang): resolve go source module dir natively on windows
- `parseGoModFile` passed the resolver's posix (`/c/...`) path to the go toolchain, so go source resolution silently found nothing on windows
- skip container-built fixtures (go archs, java builds) and golden image tarballs on windows; stereoscope's layer cache filenames contain `:`, which windows rejects
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: keep posix paths posix on windows
- `ChrootContext.ToChrootPath` nested volumes (`D:\cwd\C:\root`) when the scan root was on another drive, so the prefix was never trimmed
- bitnami, python wheel/egg, and alpm parsers used `filepath` on posix location paths
- alpm backup-file lookups silently missed on windows
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* ci: pull the test fixture cache for windows unit tests
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: don't map virtual file tree paths through the host on windows
`FiletreeResolver` without a chroot (snaps) ran request paths through `filepath.Abs`, rooting them on the current volume (`/payload.txt` -> `/d/payload.txt`), so snap scans on windows found nothing.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: match tar entry globs as posix on windows
- `matchesAnyGlob` used `doublestar.PathMatch` (host separator) on tar entry names, so tar-wrapped java archives silently produced nothing on windows
- test helpers now treat binary snippet keys and `FromFile` locations as posix
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* incorporate windows path fix
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* incorporate go-rpmdb close fix
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* close squashfs fixture writers before finalizing
- windows dir listings can report a zero size for files still open, so the manifest was packed empty
(also adds a temporary windows diagnostic for the binary cataloger)
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* temp: surface windows binary cataloger diagnostics
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: keep directory symlink targets posix on windows
- the directory indexer joined symlink targets with native separators, so links in the posix file tree pointed nowhere and globs on link names found nothing
- binary fixture logical keys used native separators, so nested snippets were picked as the fixture dir on windows
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* keep addSymlinkToIndex under the funlen limit
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* drop windows skips covered by the stereoscope and go-rpmdb bumps
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
The `perl-binary` classifier now emits `pkg:cpan/perl@<version>` instead of `pkg:generic/perl@<version>`.
`perl` is itself a CPAN distribution and carries advisories under that name, so typing it as `generic` left every one of them unreachable.
This changes an existing purl. Anything keyed on `pkg:generic/perl` (allowlists, policy, SBOM diffs) needs updating.
Builds on the cpan cataloging branch, which adds the package type this uses.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* feat: catalog CPAN distributions installed by perl clients
Adds `perl` as a language and `cpan` as a package type, with two catalogers behind them.
- `perl-cpan-installed-cataloger` reads `.meta/*/install.json` (cpanm, cpm, carton) and `auto/**/.packlist` (anything installed through `ExtUtils::MakeMaker` or `Module::Build`, including CPAN.pm). Both globs are unanchored, so a local-lib or carton application tree is found the same as a system install.
- `perl-cpan-meta-cataloger` reads an unpacked release's own `META.json` or `META.yml`, gated on a sibling `MANIFEST` so source checkouts are ignored.
Packages are keyed on the **distribution**, not the module, because that is what CPAN, MetaCPAN and the advisory data all use. `LWP.pm` belongs to `libwww-perl` and reporting it as `LWP` would make every advisory for it unreachable. The distribution name comes from `install.json`'s `dist` field, and from the `auto/` path for packlists.
purls are `pkg:cpan/<distribution>@<version>`, with the PAUSE author added as an `author` qualifier when the evidence carries it. Metadata comes in two types: `cpan-distribution` for installed evidence and `cpan-unpacked-release` for a release sitting on disk. The tiers differ in more than a name, so a consumer should not have to string-match a cataloger name to tell them apart: an unpacked release has no PAUSE path and therefore no author and no file list, and "installed and loadable by the interpreter" is a materially stronger claim than "a source tree exists here".
A packlist's version comes from `perllocal.pod`. EUMM writes it and the packlist from the same variables in the same install target, so the `auto/` path segments and the perllocal `Module` name are the same key, and the recorded `VERSION` is what was evaluated at build time rather than what can be read back statically. Scraping `$VERSION` out of the main `.pm` is the fallback, since `NO_PERLLOCAL` suppresses the file and Module::Build never writes one. Three rules pick the stanza: dashed module name, longest `installed into` libdir that prefixes the packlist path, and last match wins because the file is append-only. Without the libdir rule a second perl on the image silently supplies the version.
Where scraping is the fallback, it reads more than a plain `our $VERSION = '...'`. A version declared in the package statement (`package Foo::Bar v1.0.0;`) counts, which matters because a distribution can declare it that way and carry no `$VERSION` at all: `CPAN::02Packages::Search` is one, and without this it reports no version and matches nothing. Fully qualified `$Foo::Bar::VERSION` counts too, which is what older Dist::Zilla emitted and what real `JSON::PP` 2.27300 still carries. `qv('1.2.3')` is handled. A version computed at runtime is not, and those are reported without one rather than guessed at.
Packlist entries are parsed the way `ExtUtils::Packlist` writes them: a line can carry space-separated metadata after the path (`/path/to/File.pm type=file`), which is what `installperl` produces, so the suffix is stripped rather than the line being cut at its first space.
`META.yml` is read alongside `META.json`, because about 43% of current CPAN releases ship no `META.json` and they skew old, which is where the advisories are. Where both sit in one directory the `META.json` wins. A `META.yml` a strict parser rejects skips that directory rather than failing the scan, which is routine rather than defensive for pre-spec releases.
A packlist is literally a file list, so its paths are surfaced through `pkg.FileOwner`. They are reported as recorded, unfiltered: a path the packlist claims and the filesystem lacks means the file was removed or overwritten out from under the installer, which is worth seeing rather than hiding.
Build leftovers under `~/.cpanm/work` and `~/.cpan/build` are skipped. They genuinely are unpacked release tarballs, `MANIFEST` included, so the `MANIFEST` gate admits them and a path exclusion is the only signal available. Without it every distribution on an image that did not clean up is reported twice. The cost is that a tarball deliberately kept under `~/.cpan/build` stops being reported.
Two behaviors that look wrong but are not:
- a distribution can be reported twice at different versions. `libwww-perl` 5.836 bundles `HTTP-Date`, `HTTP-Message` and `LWP-MediaTypes`, and installing it over the modern standalone releases overwrites their `.pm` files. Both versions are genuinely present, so the merge refuses to pair disagreeing versions rather than hiding one.
- a distribution with no readable version is reported without one rather than dropped, so it stays visible.
The coverage boundary is stated in full in the `package perl` doc comment. In short: modules installed from distro packages are out of scope, since packagers strip CPAN metadata with `NO_PACKLIST` and deb, rpm and apk already report them. Core and dual-life distributions bundled with the interpreter have no coverage, because nothing on disk carries their versions; the interpreter itself is reported separately. Vendored trees, `App::FatPacker` output and PAR archives are invisible, because what they carry is module identity with no offline map to a distribution. A packlist-derived name is the installer's `NAME` and may not be the distribution name, which is left to the vulnerability data to resolve.
One correction worth calling out, since it is easy to arrive at twice: a sibling `MANIFEST` is the only thing separating an unpacked release from a source checkout. An earlier version of the guard also rejected any tree containing a `dist.ini`, on the theory that it marked a Dist::Zilla source tree. dzil ships `dist.ini` *inside* the tarballs it builds and lists it in the generated `MANIFEST`, so that exclusion was silently skipping real releases, `URI` among them.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test(cli): bump coverage image package count for the new perl fixture
The CPAN fixture added to `image-pkg-coverage` (`URI-5.35`) is shared with the CLI suite through a symlink, so the squashed package count goes from 42 to 43.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(perl): take CPAN distribution name and version from install.json dist
cpanm writes the main module's `$VERSION` into `version` and the release distvname into `dist`, and the two often disagree (`Carp-Assert-More-2.9.0` records `2.009000`). Trimming `version` off `dist` failed on those, so the whole distvname became the package name, and the module version never paired with the packlist record for the same install, which reported the distribution twice.
- name and version are now split out of `dist` at the last `-` followed by a version, as CPAN::DistnameInfo does
- a `-TRIAL` suffix is dropped, so `Try-Tiny-0.26-TRIAL` reports as `Try-Tiny` `0.26`
- the PAUSE path fallback uses the same split
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
SPDX license list 3.29.0 rewrote every opensource.org seeAlso URL from
opensource.org/licenses/<ID> to opensource.org/license/<ID> (and e.g.
opensource.org/license/mit/ to opensource.org/license/MIT). Package metadata
in the wild (Maven POM <license><url>, npm, etc.) still uses the older forms,
so regenerating the list would stop LicenseByURL from resolving them.
Fall back to a normalized index of the opensource.org entries that ignores the
www. prefix, /licenses/ vs /license/, case, and a trailing slash. The index is
built from the generated urlToLicense map, so it works with both the current
and the updated license list without changes to the generator. URL forms that
would normalize to different licenses are left out rather than guessed.
Signed-off-by: Will Murphy <willmurphyscode@users.noreply.github.com>
* fix: report scanner errors instead of returning a partial parse as complete
A line longer than `bufio.Scanner`'s 64KB buffer ends the scan with `ErrTooLong`, which only `Err()` reports. Parsers now return what they found along with the error (recorded as an unknown for the file), or log it where there is no error to return.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: allow lines up to 1MB in line-oriented package metadata parsers
Adds `internal.NewLineScanner`, a `bufio.Scanner` capped at 1MB per line rather than the 64KB default, the same cap the alpm parser already uses. Long single-line fields (descriptions, license text, dependency lists) now parse instead of ending the scan early.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: use io.ReadFull where a short read was silently accepted
- dotnet bundle 7-bit int decoding could read a zero-byte result as a zero byte
- wordpress plugin header parsing included the unread tail of its buffer
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: finish moving line scanners to the 1MB cap and checking scanner errors
- java manifest, pom.properties, Gemfile.lock, requirements.txt, homebrew formula, alpm and the purl/cpe decoders now use `internal.NewLineScanner`
- snap kernel changelog reports a scan error instead of silently dropping the base kernel package
- the attestation output reader keeps draining the pipe after a scan error so cosign can't block on a full pipe
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: stop logging the full setup.py line once per dependency match
The debug log ran once per regex match and included the whole line, so log volume grew with the square of line length. With 1MB lines that is hundreds of GB of debug output from a single crafted file.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(gentoo): don't dereference a nil location when a LICENSE scan fails
`extractLicenses` accepts a nil location, but the scan error log added earlier on this branch dereferenced it.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(gentoo): bound license group expansion on hostile license_groups
- dedup within a group uses a set instead of `slices.Contains`, which was quadratic in tokens per line and got much worse under the 1MB line cap
- each group is expanded once and memoized, so a chain of groups that reference each other repeatedly no longer expands exponentially
- repeated references to the same group within one line are skipped
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(javascript): don't panic on a yarn v1 lockfile line missing its value
A v1 `yarn.lock` dependency line without a version constraint, or a
`version`/`resolved`/`integrity` line without a value, indexed past the end of
the split line. Such dependencies are now skipped and empty fields left empty.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(alpine): don't panic on an apk db ACL or checksum line before its entry
An `M:`, `a:` or `Z:` line that comes before any directory or file entry in
`lib/apk/db/installed` indexed `files[-1]`. Such a line is now skipped.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(erlang): don't panic on a truncated rebar.lock or .app file
The term parser read past the end of the input when a file ended mid-term:
whitespace only, an unclosed list, a lone `<`, or a comment ending in `\r`.
These now return an "unexpected end of input" error.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(lua): don't panic on a truncated .rockspec
The rockspec parser read past the end of the input when a file ended mid-statement:
a lone `[`, an unclosed `[[...]]` value or `[...]` index, or a `local` assignment cut off
before its `=`, value, or line end. These now return an "unexpected end of input" error.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(swift): don't panic on a Podfile.lock pod without a version
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(gentoo): don't panic on a short obj line in portage CONTENTS
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(spdx): don't panic on an opam checksum without an algorithm
The opam cataloger stores checksums verbatim, so a malformed one reached the SPDX encoder and panicked there, outside the cataloger's recover. Checksums without an `=` are now skipped.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* refactor: use strings.Cut where a split was only indexed after a Contains check
Also guards `windows.FromPosix` against a path with no separator.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: guard untyped package metadata fields
- new `any`-typed metadata fields must be allowlisted with a reason
- allowlisted untyped fields are pushed through `SetID` and every encoder with
non-finite, deeply nested and large values
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: fuzz every generic cataloger parser
One fuzz target over every glob-selected parser in the capabilities data, run
through its cataloger task. Each input must not panic, must finish in time and
must keep allocations proportional to its size. Seeds come from each parser's
testdata plus the string literals in its package.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: scan malformed inputs under memory and time limits
A directory of small malformed files (deep nesting, oversized declared counts,
truncation, type mismatches, non-finite floats) is scanned by the CLI with a
memory limit and an RSS ceiling. The scan must exit 0 and write an SBOM.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: stop a fuzz input once its live heap passes the allocation limit
The allocation check only ran after the parser returned, so an input on its way to exhausting memory ran the machine into swap first. The harness now polls the live heap while the parser runs and panics past the same limit, which the fuzz engine saves as a crasher.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix: bound nesting depth and error context in the erlang and lua parsers
- a deeply nested rebar.lock or .rockspec overflowed the goroutine stack, which is fatal rather than a recoverable panic, so one file could take down the whole scan. Both parsers now stop at 1000 levels with an error.
- parse errors echoed whole input lines into the SBOM as unknowns, and lua appended another copy at every nesting level. `PrintError` now clips to a window around the error column, and lua decorates once at the top.
- the Podfile.lock error no longer echoes the pod entry.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: close ways the malformed-input checks could pass quietly
- a ctx-aware parser hitting the fuzz timeout now fails the input instead of racing the hard deadline
- the heap watchdog measures growth from a post-GC baseline rather than an absolute process-wide number
- a saved fuzz input naming a target that no longer exists fails instead of being skipped
- the CLI malformed-input scan fails on a recovered parser panic, not only on a non-zero exit
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Companion files next to AI models (`config.json`, `README.md`, and the OCI model.file, license and model config layers) are now read with named size limits, and are only used when they look like what they claim to be.
- JSON companions are read whole up to a limit and skipped when larger, rather than being fully buffered or parsed truncated. OCI companion layers are fetched one byte past the cap so oversize is detectable.
- README and license layers read a prefix. The frontmatter block is capped at 256 KB and list values at 32.
- dir and OCI scans share one 8 MB safetensors header cap. Directory scans used to allow 100 MB.
- `config.json` is only used when it has `architectures` or `_name_or_path`, and is looked up beside the model or in its parent directory, no longer all the way up to the scan root
- OCI model.file layers are processed in a stable order, so the chosen name no longer changes between runs
- README frontmatter:
- a list-form `license:` is accepted and no longer drops `base_model`
- `null` values are ignored
- delimiters must be exact `---` lines
- model names from `_name_or_path` / `base_model`:
- Windows paths are handled
- `.`, `..` and `/` fall back to the next name source
- names are capped at 256 characters
Behavior changes worth noting:
- a `config.json` two or more directories above the model no longer names it
- JSON `config.json` files without `architectures` or `_name_or_path` (e.g. `generation_config.json`, app configs) are ignored
- safetensors headers over 8 MB in directory scans are now rejected, matching OCI scans (real shard headers are well under 1 MB)
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(generic): recover from parser and processor panics
- a panic while parsing one file becomes an unknown for that location instead of failing the whole SBOM
- a panic in a package processor is logged as a warning and the unprocessed results are kept
- stacks go to the trace log, not into the unknown text
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(unknown): keep errors that unwrap to nil
`sync.PanicError` with a non-error panic value (e.g. `panic("boom")`) unwraps to nil, and was being dropped entirely from the error graph instead of being kept as a leaf.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(task): don't put the stack trace in task panic errors
Task panics are still fatal, but the error now reads `panic in task "<name>": <value>` and the stack goes to the debug log.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test(generic): cover chained processors when a later one panics
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(file): recover from panics in the executable and file digest catalogers
A panic while reading one file is now recorded as an unknown for that file instead of failing the whole SBOM, the same as generic catalogers.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(ai): use go.yaml.in/yaml/v3 to parse model card frontmatter
gopkg.in/yaml.v3 panics on some malformed merge-key payloads (for
example "<<:\r? 0:"), which crashed the gguf and safetensors catalogers
when a model README contained such frontmatter. Switch the Hugging Face
frontmatter parser to the maintained go.yaml.in/yaml/v3 fork, which is
already used by the other catalogers and returns an error instead of
panicking. The existing error path logs and skips the frontmatter.
Fixes#5327
Signed-off-by: tk1475 <118763009+tk1475@users.noreply.github.com>
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
* chore: finish moving off gopkg.in/yaml.v3
- move the snap catalogers and `internal/capabilities` to `go.yaml.in/yaml/v3`. The snap parsers decode yaml from scanned sources and hit the same merge-key panic as #5327
- enable `depguard` with a deny rule for `gopkg.in/yaml.v3` so it can't be reintroduced
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: tk1475 <118763009+tk1475@users.noreply.github.com>
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: Alex Goodman <wagoodman@users.noreply.github.com>
the index was ported from a prototype with per-node `RWMutex` lock upgrading and JSON marshalling. Its only consumer (`archive.Resolver`) builds it sequentially during extraction and then only reads it, so neither was used. `lock.go` also compared method-value pointers to tell the read lock from the write lock, which can't distinguish lock instances.
removed the locks, the upgrade and retry paths that existed only for concurrent writers, and the JSON methods and their test. The radix logic is unchanged, and the README now points at `archive.Resolver` rather than a nonexistent `archive.Index`.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
SPDX wrote `FileName` from `RealPath` only, so two archives' `META-INF/MANIFEST.MF` became two SPDX files with the same name and nothing saying which archive each came from. Worse, a consumer reading just the name would take a nested file for one at that path in the scan root.
- a file inside an archive is now named by its full chain, using the same `:` convention and escaping as `ArchivePath` / java's `VirtualPath`: `app.war:WEB-INF/lib/a.jar:META-INF/MANIFEST.MF`
- `FileComment` gains an `archivePath: ...` line (next to the existing `layerID: ...`, one per line), so decoding recovers `ArchivePath` and the entry's own `RealPath` without guessing where the chain ends
- files outside archives are unchanged
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
the legacy recursion links a nested jar to the jar that holds it: a `DEPENDENCY_OF` edge from inner to outer, and `JavaArchive.Parent` on the inner package. Inside an extracted archive java called `parse(ctx, nil)`, so both went away as soon as the feature was on. What replaced them was `file CONTAINS package`, which SPDX and CycloneDX drop, so those formats lost all war-to-jar structure.
- `archive.Traversal` gains `Parent` and `Packages` (the packages found describing that archive, filled in by the task after its sub-pipeline runs)
- java picks the enclosing jar's main package from the parent traversal (the one whose virtual path is the archive's own; pom-derived aux packages carry a suffixed one) and passes it to `parse`, then sets `Parent` the same way the legacy path does
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
with nested cataloging on, java leaves root-level archives to the archive task, so a root `app.jar` package only exists once that task has run. With the default `FilesOwnedByPackageSelection` the root file catalogers ran in an earlier group, never saw the package, and `app.jar` got no file metadata or digests. It did get them with the feature off.
- in owned-by-package mode the groups are now packages, then the archive task, then the root file tasks. Other modes are unchanged
- the file tasks merge their results into the SBOM instead of replacing the maps, since the archive task has already written nested files into them
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
with `FilesOwnedByPackageSelection` the file catalogers take every package location in the SBOM and look it up by `RealPath` in the resolver they were given. Inside an archive, the java package's location is the archive itself *in its parent's frame* (`WEB-INF/lib/dep.jar`), so the sub-pipeline looked that path up inside `dep.jar`'s own resolver: a miss, or an unrelated same-named entry by coincidence. At the root, coordinates from inside archives were looked up against the scan root the same way.
`coordinatesForSelection` now keeps only coordinates whose `ArchivePath` matches the current frame (the traversal's `ContentsArchivePath()`, or blank at the root).
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
a read error while copying one entry (a zip CRC mismatch, a corrupt deflate stream) was returned from `add`. That aborted the walk, and `Extract` threw away every entry already stored. So a large war with one damaged, irrelevant entry produced zero packages with the feature on, while java's legacy path (which reads only the entries it needs) still cataloged it.
- a read error on one entry now drops that entry and marks the resolver truncated (`entry "x" could not be read: ...`), and the walk continues
- if the stream itself breaks partway (a truncated `.tar.gz`), whatever was read before the break is kept and the archive is truncated. With nothing read it is still an error, as before
- disk limit and decompression budget errors still stop the walk
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
the archive's `CONTAINS` edges were built from every coordinate in the sub-pipeline's scratch SBOM, which gets one of two things wrong:
- with the default `FilesOwnedByPackageSelection`, nothing records the nested archives as files, so the archive-to-archive chain was missing and grandchild packages hung off coordinates with no incoming edge
- the scratch SBOM also holds coordinates from the parent's frame, most notably unknowns recorded at the archive's own coordinates (e.g. a jar with no manifest). Those turned into `x.jar CONTAINS x.jar`
edges now go to the nested archives found in it (always, whatever the file selection), plus coordinates whose `ArchivePath` is this archive's own chain, i.e. files that are actually inside it. The walk reuses that one discovery pass for the recursion.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
with nested cataloging on, java stops recursing into jars itself (the task owns recursion) and its own "nested archives not cataloged" unknown is suppressed. The task then just returned at the depth bound. So `MaxDepth: 1` on a war silently found fewer jars than leaving the feature off, and nothing in the SBOM said so.
at the bound, every archive that sits there (by content sniff) is now recorded as a "nested archive not cataloged: the nested archive depth limit was reached" unknown. The CLI description for `nested-archive-max-depth` notes that it replaces java's unbounded jar-in-jar recursion.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
the limiter bounds what's held *at once*, and each archive gives its bytes back when it's released. So a recursive bomb (42.zip shape) never trips a limit: every level fits on its own, and the scan just runs for effectively forever. Capping each archive at N× its own size wouldn't help either, since every level of such a bomb is high-ratio, so the work still grows about N-fold per level.
- each top-level archive gets one decompression budget of `max(100 × its size, 1 GiB)`, spent by every archive nested under it at any depth
- the budget is attached to the context when the walk enters a top-level archive, so `Extract` picks it up without a signature change
- once it's spent, the archive being extracted is truncated with a "possible decompression bomb" unknown and the walk moves on. The next top-level archive starts fresh
real trees (a war of jars, a layer of tarballs) expand a few times over, well under the ratio.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
`put` had no per-entry cap, unlike the legacy java path which capped every file at 2 GB through `intFile.SafeCopy`. At depth 1, one entry of an overlapping-entry zip bomb could decompress all the way to the 100 GB disk limit before anything stopped it.
- entries are now read through the same 2 GB cap (`intFile.PerFileReadLimit`, now exported). An entry over it is dropped after cap+1 bytes and the rest of the archive is still extracted
- `Resolver` gains `TruncatedReason`, so the truncation unknown says what happened ("an entry larger than 2.0 GiB was skipped", or the existing disk-limit text) instead of always blaming the limits
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
a negative `MaxDepth` really was unbounded, and nothing noticed an archive nested inside a copy of itself. A zip quine at `-1` recursed until memory, disk or file descriptors ran out, running the full sub-pipeline at every level.
- the walk keeps a stack of the digests of the archives it's inside, and an archive identical to one of them is recorded as an unknown instead of being entered
- negative depth now stops at 16 levels (docs and CLI description updated)
- cancellation is checked between candidates, so a cancelled scan stops walking instead of finishing the archive tree
the three unknown-recording blocks in `processArchive` became a one-line helper to stay under `funlen`.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
`identifyFormat` accepted any extractor mholt identified, which put the 7z and rar decoders (`bodgit/sevenzip`, `rardecode`) on untrusted input with no test coverage. Only zip, tar and compressed tar are opened now, until other formats are deliberately supported.
This also fixes java resource adapters: a `.rar` that is really a zip used to be identified by name as RAR and fail to extract. An unsupported identification now falls through to the zip end-of-central-directory check, so it's read as the zip it is.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
sub-pipeline catalogers were already run under `runTaskSafely`, but extraction wasn't: the zip, tar and compressor decoders run on untrusted bytes, and a panic in one unwound the whole walk and failed `CreateSBOM` with "failed to run tasks".
- `processArchive` recovers and returns the panic (with its stack) as an error at the archive's coordinates, so it becomes an unknown and the sibling archives are still cataloged
- `Extract` now defers `Cleanup` on failure instead of calling it on each error path, so a panic no longer leaks the limiter charge or the spill file
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
each `ArchivePath` segment came from `loc.Path()`, so it was the link path when an archive was reached through a symlink and the real path otherwise. That depended on how discovery happened to find the archive. The package location said `store/3f9a1c` while the files inside the same jar were recorded under `...:app/lib/json-simple-1.1.1.jar`, so a consumer couldn't join a package to its files by prefix.
The chain for extracted files is now built from `RealPath`. `Traversal.Location` still carries the access path, so java's name check and `VirtualPath` metadata behave the same as before. The field comment and schema now say every segment has links resolved.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
`discoverArchives` collects MIME hits first, and those are bare real paths. When the glob pass then found the same file through a symlink, it was dropped as a duplicate along with its access path. Java decides whether to describe an archive by its name, so a content-addressed jar (`store/3f9a1c`, linked from `lib/json-simple-1.1.1.jar`) was extracted but produced no package. This is the pnpm / nix / bazel layout.
When deduping, the candidate now takes the first location that carries an access path.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
`:` was escaped as `%3A`, but `%` itself wasn't, so an entry named `a:b.jar` and one named `a%3Ab.jar` got the same `ArchivePath`. Their contents then collide in `FileMetadata` / `FileDigests` (last writer wins) and share artifact IDs.
`%` is now escaped as `%25` before `:`, which makes the encoding injective. The encoding is documented on `Coordinates.ArchivePath` and in the schema.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
`ArchiveSearchConfig{MaxDepth: 2}` is the natural literal, but zero memory and zero disk meant "forbid both", so every archive was truncated to nothing and only unknowns came out.
- `MaxMemoryBytes` / `MaxDiskBytes` of `0` now mean the default, negative is still unbounded
- the limiter keeps its own zero-forbids semantics internally; it's just no longer reachable from config
- tests that relied on zero to force spilling now pass `archive.Limits` directly, and the two "both limits zero" tests are gone since that config no longer exists
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
matches how the per-cataloger progress rows are titled (`Default` only), so the archive row reads like its siblings.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* build(deps): bump go-make to v0.8.1
Picks up golangci-lint 2.13.2 (built with go1.27) in place of 2.11.4 (go1.26).
The old build panics outright on a go1.27 GOROOT with "file requires newer Go
version go1.27", and an older mise-shim build fails the quieter way: it loses
type info and reports zero findings with exit 0, so a ruleguard run comes back
green without having evaluated anything.
Also folds in what the newer linter flags:
- `reflect.Ptr` -> `reflect.Pointer` in three files, applied by --fix
- `tw.CellFormatting{MergeMode: ...}` -> `tw.CellMerging{Mode: ...}` in the
cataloger info table, for a staticcheck SA1019 deprecation
Note the deprecation text names `CellConfig.CellMerging.Mode`, but the field is
`CellConfig.Merging`; `CellMerging` is the type. That command path has no test
coverage, so the swap was checked by rendering the same hierarchical-merge table
both ways and diffing: byte identical.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* chore(lint): tune goconst for data tables
goconst got stricter in golangci-lint 2.13 and now counts strings in map
literals and struct fields, which lights up every lookup table in the repo. It
was already enabled, so this is main going red on a linter bump rather than
anything new in the tree: 101 findings, and the default threshold of 3 means a
vendor name appearing in three rows of a table reads as a magic constant.
Raising `min-occurrences` to 5 takes that to 14. The rest split by whether the
repeated string is data or an actual constant:
- `cpegenerate` is excluded by path. It maps artifact names to vendors,
products, and maven group ids, and repeats strings 35, 83, even 293 times, so
no threshold reaches it. Hoisting those would trade a readable table for
indirection.
- the two SBOM fixture builders in `format/internal/testutil` get a
function-level nolint each. The repeated names and versions are what the
golden files assert against.
- `mysql-cluster` and `php` were genuine repeated literals and are now
constants.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* chore(test): extract measureAlloc into a shared helper
The allocation-budget tests added during the unbounded allocation work grew a
copy of the same helper in every package they touched: four identical
`measureAlloc` definitions plus two more spelled out inline against
`runtime.MemStats`.
Now `testutils.MeasureAlloc`, at the repo root so `internal/spillbuf` can reach
it too. The doc comment carries the two things that were only recorded in some
of the copies: TotalAlloc is process-wide so callers must not use t.Parallel,
and a budget assertion only means something if the fixture would really blow
past it unguarded, so measure the unguarded path as well.
No behavior change, the body was identical in all six.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* chore(lint): add ruleguard rules for overflow and swallowed EOF
Two rules out of the recent unbounded allocation work (#5187, #5195, #5267,
#5291, #5292, #5293, #5294).
`noOverflowingBoundsCheck` flags `offset + size > limit` (and <, <=, >=) where
an operand is unsigned, neither addend is constant, and the operands are at
least 32 bits. Unsigned addition wraps, so an offset out of a file header can
carry the sum past the wrap point and pass a check it should have failed. The
fix is to subtract against the limit instead. Constant addends are excluded
because `pos+4 <= uint32(len(buf))` is the normal way to step a cursor, and the
32-bit floor drops nibbles widened for a comparison. Would have caught four
sites across #5195 and #5291.
`noSwallowedEOF` flags `return nil` / `return nil, nil` on an io.EOF check. The
caller gets a zero value it cannot tell from real data, and anything decoding a
fixed-size struct off it parses the padding as content. That is the shape behind
#5186.
Both decide from the expression alone, and writing the fix deletes the pattern,
so neither needs a suppression anywhere in the tree. That was the bar: rules for
the allocation and decompression classes were tried and dropped, because the
bound they look for lives in a preceding statement (or, for a decompressor
constructor, always downstream), so no correct implementation can silence them.
That would have meant ~15 permanent nolints on code that is already right. The
rules here that do work don't judge whether a bound exists, they route to a
wrapper that takes one: noDirectTempFiles, noDirectELFOpen.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: pin allocation budgets on the alpm, deb, and changelog bounds
The bounds added for #5292, #5293 and #5294 are asserted by the error they
return. That keeps passing if the bound is removed and something else rejects
the input later, which is the failure these were written for. Each of these
measures bytes instead, and proves the fixture is really a bomb first so a
fixture that quietly stops being one fails rather than passing.
- alpm mtree: 16MB of newlines, well inside the 64MB byte cap, costs 16.5GB
uncapped against 238MB with maxMtreeLines. That 16.5GB is the measurement the
line cap was added for and it was recorded only in the commit message. Note
what the cap actually buys: 300k entries still cost a few hundred MB, per
concurrent cataloger.
- deb members: 1.28GB against 392KB. Both arms read with io.ReadAll on purpose,
since the stream itself never holds much whatever the cap says; what the cap
protects is a consumer that buffers, and the real one is a tar reader building
entries.
- snap changelog: 1.28GB against 175KB. This one has no explicit cap at all. It
is held by bufio.Scanner refusing a token past MaxScanTokenSize, which is
incidental, so switching to io.ReadAll or a bufio.Reader to pick up long lines
would remove it silently.
Nothing added for the kernel module reader: its copy is already
io.Copy(tempFile, io.LimitReader(rc, max+1)), so peak disk is bounded by
construction and heap is streaming. A heap budget would assert the wrong
resource, and the three existing subtests cover the cap, at-cap fidelity, and
spill cleanup.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: pin allocation budgets on the ai and alpm byte caps
Follows the same pattern as the alpm line cap: measure bytes, and prove the
fixture is really a bomb first so one that quietly stops being a bomb fails
rather than passes.
- safetensors: 252MB against 2.6KB. The existing over-cap case deliberately
omits the body, since the guard rejects on the declared length before reading
anything. That is the right test for the guard but it would keep passing if
the check moved after the read, so this one delivers the bytes.
- gguf: 268MB against 168MB. copyHeader io.Copy's whatever it is handed, so
maxHeaderSize holds only because every caller wraps the reader first. This
pins that the limit is where the bound lives, so dropping the wrapper at a
call site fails here.
- alpm byte cap: 740MB against 132MB, the counterpart to the line cap budget.
Worth knowing for anyone writing the next one of these: a single enormous line
does not exercise the alpm byte cap. go-mtree runs its own bufio.Scanner and
refuses a token past 64KB, so that fixture fails as "token too long" before the
byte cap is consulted. The fixture here uses ordinary 1KB lines, enough of them
to pass 64MB while staying under the 300k line cap, so the byte cap is actually
the thing under test.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* test: assert bytes retained instead of allocated for gguf copyHeader
the bounded arm writes into a bytes.Buffer, whose doubling growth costs a
multiple of what it holds and lands on a different power of two depending on
the platform. it measured 168MB locally and 320MB on CI, past the budget.
buf.Len() states the same property deterministically.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(golang): bound UPX decompression by what the input can justify
A UPX block's `sz_unc` was passed straight to `make([]byte, n)`, so four bytes of
attacker input spanned the full uint32 range. A ~300KB file could drive multi-GB
resident memory, which Go reports as a fatal runtime OOM that no `recover` on this
path can contain. Syft parses binaries out of arbitrary images, so that is reachable
from any scan.
Two bounds do the work, and UPX's own invariants make them exact:
- all blocks together reconstruct `p_filesize`, so a running remainder caps the sum.
Bounding only per-block would leave the total at (block count x original size), so
the remainder is the part that actually closes it
- `p_filesize` sizes the output buffer, so it is capped both absolutely and against
the size of the file on disk. The absolute cap alone left a ~60 byte header able to
claim 500MB; the ratio alone cannot work either, since LZMA encodes a run of N
equal bytes in O(log N) and a `go:embed` of 120MB of zeros legitimately packs 209x
The output buffer is also allocated on the first block that survives validation
rather than up front, so a header claiming a large size with nothing decodable behind
it costs nothing.
Alongside that, several ways the block loop could be steered off its own buffer:
- `parseELFPTLoadOffsets` bounds-checked program headers with `phStart+phentsize >
len(buf)`, which overflows for a `p_offset` near 2^64 and lets an out-of-range entry
through into a slice index. Switched to the subtraction form, and a `phentsize`
shorter than an ELF64 program header is rejected since the reads use fixed offsets
- block placement had the same overflow shape, and on failure it skipped the copy and
fell through. `outputOffset` derives from the rejected offset, so a `p_offset` at the
uint64 ceiling wrapped it to 0 and the next block landed on the reconstructed ELF
header, still returning success. Out-of-range placement now ends the block chain,
keeping the blocks already placed since those often carry `.go.buildinfo`
- the UPX 2-byte LZMA header carries `lc`/`lp` as nibbles and `pb` as three bits, so
they can hold values LZMA does not permit. Folded into the props byte they wrap
(`lc=15, lp=15, pb=7` gives 465, which truncates to 209) and the stream mis-decodes
instead of failing
- blocks decode directly into a slice of the output buffer instead of a per-block
buffer that is then copied, so one block no longer doubles peak memory
- the block count is capped: the size budget alone still allows millions of 12-byte
blocks, each spinning up an LZMA reader
Verified against the `image-small-upx` fixture, a real `upx --best --lzma` binary:
four blocks, max `sz_unc` equal to `p_blocksize` exactly, blocks summing to 0.68 of
`p_filesize`, and a 1.96x expansion against the packed file, so every bound holds with
headroom on genuine output.
Note `p_blocksize` is deliberately not used as a ceiling on `sz_unc`. It adds nothing
the remainder does not already cover, and real output sits at exactly `p_blocksize`,
so it would run with zero headroom against a value the format does not promise.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(golang): report UPX binaries we cannot unpack as unknowns
A packed Go binary we fail to unpack means its packages are silently missing from the
SBOM. `getBuildInfo` checked the decompression error for nil and discarded it, so a
rejected file surfaced only the original `buildinfo.Read` error, and that one is
deliberately silenced because it is usually just "not a Go binary".
`decompressUPX` now marks the cases worth reporting with `errUPXDecompress`: it got
past the header and the method dispatch and still could not unpack the file. Everything
else stays quiet, which matters more than it sounds. `upx` defaults to NRV2B unless
`--lzma` is passed and only LZMA is implemented here, so treating any decompression
failure as reportable would attach a golang-cataloger unknown to every packed non-Go
binary in an image. Making the reportable case opt in rather than the quiet case a
growing list also means a guard added later is silent by default.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* refactor(golang): drop the vendored xcoff parser
The only thing the golang cataloger read from it was `TargetMachine`, which is the same
two magic bytes `getGOARCHFromBin` had already matched on to dispatch there. So a full
XCOFF walk (string table, symbol table, every relocation table) ran to recover a value
that was already in hand.
`getGOARCHFromBin` reads those two bytes directly now. That is also strictly more
correct: `xcoff.NewFile` bailed with "no symbol table" when `symptr == 0`, so a stripped
XCOFF binary reported no arch at all.
Fixtures move to `testdata/xcoff/`, since they are no longer a package's own testdata.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* feat(internal/file): report a reader's real size behind one helper
Bounds get written as "does this claim exceed what the reader actually holds", and the
subtlety is that a reader's own answer cannot be trusted: an `io.SectionReader` reports
the nominal length it was built with, so one constructed with `1<<63-1` will happily
claim to hold 8EB. Confirming the last byte is readable is what separates a real size
from a nominal one.
Returns `(int64, bool)` rather than a bare size, since there are five ways to not know
and collapsing them to 0 makes every caller reinvent "0 means stop bounding".
The doc also names the wrapper trap: a type embedding `io.ReaderAt` as an interface
promotes only `ReadAt`, so it answers no size at all however large the reader beneath it
is, and a bound written against it silently does nothing.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(golang): bound every section a reader can expand, not just the name table
`CheckSectionNameTable` bounds the one section `elf.NewFile` expands while parsing, which
is enough for a caller that only parses. It is not enough for one that goes on to read
sections: those are expanded lazily, so a 260KB ELF declaring a compressed `.symtab`
drove 1.3GB of allocation through `goversion`, which opens the file with `debug/elf`
inside its own package and cannot be routed through `elfutil.NewFile`.
`CheckAllSections` is that gate, and is deliberately named as the superset so the
relationship to the narrow one is structural rather than documented.
Also exports `ErrDeclaredSizeExceeded`. A refusal here costs the SBOM a package, so
callers report it as an unknown, and matching on an error string to decide that is not
something to build a reporting policy on.
A ruleguard rule now flags `buildinfo.Read` and `version.ReadExeFromReader` outside
`scan_binary.go`. Both open `debug/elf` internally and are only bounded there by the
wrappers that gate the reader first, which the existing rule on `elf.NewFile` could not
see through.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* feat(internal/spillbuf): sparse offset-addressed buffer that spills to disk
For output that arrives out of order, at offsets the input itself declares: a
decompressor placing extents, an archive rebuilding a file from chunks. A plain `[]byte`
cannot do that without either pre-sizing to a length the input claims or growing to the
furthest offset it names, and both hand a hostile input an allocation knob.
The load-bearing property is that it reports only what it actually stored. `Size` is the
contiguous run written from offset zero and a read past it is `io.EOF`, never the zeros
an unwritten region would hand back for free. Sparse storage serving holes as real
content is what lets a write offset stand in for output nothing produced, so anything
sizing an allocation against this reader (`saferio` in `debug/elf` does exactly that)
stays bounded by work actually done.
`FirstGap` answers where the next write fits, so callers filling holes do not need the
extent bookkeeping and cannot re-derive that rule wrongly, and it only reports gaps
lying entirely inside `[0, within)`. The extent type is unexported for the same reason.
The rest of the contract worth stating: reads and writes after `Close` fail with
`os.ErrClosed` rather than looking like an empty buffer, and a write that fails leaves
the buffer unchanged.
Memory is bounded by the limit rather than by how much is written: filling 1MB and
filling 64MB cost about the same. What lands on disk past that limit is the caller's
bound to set, not this package's. Growth is geometric and capped, because sizing the
tier to exactly what each write needs is quadratic once a caller streams through it in
small chunks.
The golang cataloger is the only consumer today, and nothing else in the tree has this
shape.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* fix(golang): rebuild UPX binaries through spillbuf and read the reconstruction everywhere
Two amplification paths were still reachable. A 128 byte input allocated 16MB by
trusting `sz_cpr` before weighing it against the bytes the file actually holds, and the
reconstruction was assembled in a `[]byte` sized from the header's declared original
size, so a 301KB input allocated 300MB.
The reconstruction is a `spillbuf.Buffer` now, which is what makes the second one go
away: blocks are placed at the offsets the file declares, nothing is pre-sized from a
claim, and the reader's length is the contiguous run actually rebuilt. That last part is
the invariant everything downstream leans on. A reader reporting the declared size would
serve the holes between placed blocks as zeros it never stored, and a 64 byte block
parked at `p_filesize-64` then made a 300KB input hand out a 64MB reader: measured, that
drove 269MB through `getBuildInfo` and returned nothing, silently.
Placement and storage are separate concerns. `upx.go` places blocks at the offsets the
file declares and asks the buffer how much was rebuilt and where the next gap is; it
knows nothing about temp files or memory limits.
The remaining bounds are ratios against the input wherever they can be, since an
absolute ceiling only ever drops a large legitimate binary. The one absolute left is on
`p_filesize`, because padding an input is nearly free, and the LZMA dictionary is paid
for in its own block's `sz_cpr`.
Interface change: `scanFile` takes a context and hands back the reconstruction for each
binary that turned out to be packed, and the caller owns it. Everything after the build
info read (crypto settings, arch, symbols, the version scan) reads it, so a packed
binary no longer reports its packages with no symbols at all.
Also narrows what gets reported. This cataloger runs on every executable in an image, so
reporting every `getBuildInfo` error attached an unknown to every corrupt ELF, odd PE and
non-Go Mach-O slice in it. Now only four gaps, the ones that cost the SBOM something
syft could have had: a packed file we could not decode (`errUPXDecompress`), a header
the size bounds refused (`errUPXSizeRefused`), a reconstruction that came up short
(`errUPXPartial`), and an ELF expansion `elfutil` declined
(`ErrDeclaredSizeExceeded`). An unimplemented UPX method stays quiet on purpose: `upx`
defaults to NRV2B and only LZMA is implemented here, so reporting it would attach an
unknown to most packed non-Go binaries in an image.
That filtering is a new pattern. No other cataloger picks which of its errors become
unknowns, and the reason this one does is that it runs on every executable in an image
rather than on files a glob already selected.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* chore: ignore local spec/ planning notes
`/specs` was already ignored; `/spec` is the same thing under the singular name and was
not, so working notes landed in commits.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* refactor(golang): name what the unpack hands back, not where it stores it
the scan's signatures said `*spillbuf.Buffer` all the way down, which told
every reader below the scan that it cares whether the reconstruction lives in
memory or on disk. it does not. two interfaces instead, each naming only what
its side actually uses:
- `unpackedContents` for the consumers: `ReaderAt`, `Closer`, and a `Size` that
is the contiguous run rebuilt from zero
- `blockSink` for the decoder, which additionally needs `WriterAt` and
`FirstGap` to place blocks
this costs one thing worth flagging. `Close` on a nil `*spillbuf.Buffer` was
nil-safe, so callers just deferred it; widened into an interface, a nil pointer
becomes a non-nil interface and the same call panics. `closeUnpacked` absorbs
that, and it is now the only way an owner releases contents. `readerFor`,
`seekerFor` and `closeUnpacked` are the three places that resolve the nil.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* docs(elfutil): say that CheckAllSections is the decompression-bomb gate
the doc led with "bounds every section syft can drive debug/elf into
decompressing", which describes the mechanism without ever naming what it is
defending against. a reader landing on the call site could not tell that this
is the decompression-bomb check.
state the attack instead: a compressed section header declares its own
decompressed size, debug/elf believes that number and allocates it on open, and
nothing forces it to match what the compressed bytes actually yield. the 260KB
ELF that drove 1.3GB of allocation now lives in the doc rather than only at the
one call site that happened to mention it.
comments only.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* docs(golang): move the overflow note onto the bounds test it describes
the "subtraction form" note sat above the phStart assignment with ten lines of
wrap analysis between it and the `if` it was talking about, so the subtraction
it names reads as missing. it is `hdrLen-phStart < phentsize`.
move it onto that test and name the form it is avoiding (`phStart+phentsize >
hdrLen`, which can overflow past the buffer end) so the referent is not several
paragraphs away.
comments only.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* chore(golang): drop the internal/ README left over from xcoff
that README existed to record where the vendored xcoff package came from and
that Go keeps it internal, which is a provenance note worth having while
third-party code sat there. dropping xcoff took the reason with it, and
repurposing it to point at gotestdata just made it redundant: gotestdata has
its own README covering why it is under internal/, why it is not testdata/, and
what belongs in each.
internal/ now holds one self-documenting directory.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* feat(golang): add extended-stdlib scope and include patterns for symbol capture
`golang.capture-symbols` decides how much symbol data lands in the SBOM for grype's reachability analysis. It's `none`, `stdlib`, or `all` today, and the useful middle is missing: `stdlib` stops at the standard library, `all` multiplies SBOM size.
A new `extended-stdlib` configurable covers stdlib plus everything under `golang.org/x/`:
```yaml
golang:
capture-symbols: extended-stdlib
```
Also, a new `capture-symbols-include` configurable for modules that are noisy in your binaries but not everyone's. It's unioned with whatever the scope selects, so it only ever widens:
```yaml
golang:
capture-symbols: extended-stdlib
capture-symbols-include:
- github.com/klauspost/**
```
Patterns are standard doublestar globs, which matters because module paths carry `/v2`-style suffixes:
```yaml
golang:
capture-symbols-include:
- github.com/klauspost/* # compress, but not compress/v2
- github.com/klauspost/** # both
- k8s.io/client-go # exact match only
```
Ordering is `none` < `stdlib` < `extended-stdlib` < `all`. The existing three values
and the `none` default are unchanged, and the include list is inert under `none`.
Presets compile into glob lists internally, so a single matcher answers "does this
module get symbols" instead of a preset branch sitting next to a separate glob branch.
An unrecognized `capture-symbols` value still falls back to `none`, but warns now
instead of doing it silently. A malformed include pattern warns and gets skipped.
One thing worth a look beyond the feature: the `Symbols` field description in the JSON
schema was wrong after this (it claimed only `all` and `stdlib` populate anything), and
that description lives in the already-published `16.1.10`. Rather than bump a version for
a sentence, `16.1.10` is amended in place and `schema/json/README.md` grows an explicit
exception for description-only changes: descriptions only, no shape change of any kind,
`$id` unchanged. Anything else still needs a bump. Happy to split that into its own PR if
you'd rather review the policy separately.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* refactor(golang): rename capture-symbols-include to capture-symbols-modules
The key's entries are go module paths, and `-include` sitting next to `capture-symbols` reads as plausibly taking symbol or package names instead. Those spellings parse and match nothing, which is quieter than the confusion `-include` was picked to avoid, so the name now says what the list holds.
`golang.CatalogerConfig.CaptureSymbolsModules` and `WithCaptureSymbolsModules` rename with it. Nothing behavioral changes; the key is new in this PR so there is no compatibility surface.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
* feat(golang): match capture-symbols-modules across major version suffixes
`github.com/anchore/*` covered `github.com/anchore/syft` and silently stopped covering it the day it became `github.com/anchore/syft/v2`. The config keeps parsing, nothing warns, and symbols quietly go missing from the SBOM. Exact paths had the same hole: `github.com/klauspost/compress` did not cover `compress/v2` either, so no spelling short of `**` survived a major bump.
A major version suffix is part of a module's path but not part of its identity, so patterns are now matched against the module path both with and without it, using `module.SplitPathVersion` from `golang.org/x/mod` (already a direct dep, already used in this package for `PseudoVersion`).
```yaml
golang:
capture-symbols-modules:
- github.com/klauspost/* # compress and compress/v2
- github.com/klauspost/compress # same module at every major version
- github.com/klauspost/compress/v2 # v2 alone
```
Only a trailing suffix is a version, which is Go's own rule. In `github.com/anchore/syft/v2/thing` the `v2` is an ordinary path element naming a major subdirectory a nested module lives in, so it stays literal and `github.com/anchore/**/thing` is how you reach it. `/v0` and `/v1` are not valid suffixes and are left alone.
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
---------
Signed-off-by: Alex Goodman <wagoodman@users.noreply.github.com>
Keith Zantow
Keith Zantow
MetalandAlex Goodman
anchore-oss-update-botandanchore-oss-update-bot
Will Murphy
Sverrir Sigurdsson
Christopher Angelo PhillipsandAlex Goodman