Keeping the original lexical sorting of Go packages would be very hard,
as a Go program may import an unknown number of Go packages,
and we load and obfuscate one package at a time by design.
One option would be to load all packages upfront when obfuscating
main packages, but that would break the per-package caching of
ofbuscated Go packages, causing a huge slow-down in builds.
Another option would be to not obfuscate import paths,
which would clearly cause a worsening of the obfuscation quality.
The third option is to not attempt to keep the original order,
and document that as a caveat in the README.
I suspect the vast majority of Go projects won't be affected by this,
and those few that might be can always use imports to enforce the order.
Closes#693, per the decision above to not change what we do.
This type of question has popped up a number of times before,
particularly when someone unfamiliar with obfuscators in other languages
first hears about the existence of one for Go.
This means we now have a unified cache directory for garble,
which is now documented in the README.
I considered using the same hash-based cache used for pkgCache,
but decided against it since that cache implementation only stores
regular files without any executable bits set.
We could force the cache package to do what we want here,
but I'm leaning against it for now given that single files work OK.
A user correctly points out that some sentences were confusing, like:
It can also make reverse-engineering harder, as an end user could
guess what version of Go or garble you're using.
Rewrite the whole section. It now also explains what I meant by that
sentence with a practical example.
I also merged the bits about "provide your own seed" and "rotating the
seeds", because they're both talking about the same mechanism.
Fixes#666.
On my machine, garble itself goes from about 5.7MiB to about 4.8 MiB
when being built with `garble build` and `garble build -tiny`,
which is roughly a 16% decrease in size.
The 2-5% figure is dated at this point, as we do better than that now.
This value is hard-coded in the linker and written in a header.
We could rewrite the final binary, like we used to do with import paths,
but that would require once again maintaining libraries to do so.
Instead, we're now modifying the linker to do what we want.
It's not particularly hard, as every Go install has its source code,
and rebuilding a slightly modified linker only takes a few seconds at most.
Thanks to `go build -overlay`, we only need to copy the files we modify,
and right now we're just modifying one file in the toolchain.
We use a git patch, as the change is fairly static and small,
and the patch is easier to understand and maintain.
The other side of this change is in the runtime,
as it also hard-codes the magic value when loading information.
We modify the code via syntax trees in that case, like `-tiny` does,
because the change is tiny (one literal) and the affected lines of code
are modified regularly between major Go releases.
Since rebuilding a slightly modified linker can take a few seconds,
and Go's build cache does not cache linked binaries,
we keep our own cached version of the rebuilt binary in `os.UserCacheDir`.
The feature isn't perfect, and will be improved in the future.
See the TODOs about the added dependency on `git`,
or how we are currently only able to cache one linker binary at once.
Fixes#622.
We can drop the code that kicked in when GOGARBLE was empty.
We can also add the value in addGarbleToHash unconditionally,
as we never allow it to be empty.
In the tests, remove all GOGARBLE lines where it just meant "obfuscate
everything" or "obfuscate the entire main module".
cgo.txtar had "obfuscate everything" as a separate step,
so remove it entirely.
linkname.txtar started failing because the imported package did not
import strings, so listPackage errored out. This wasn't a problem when
strings itself wasn't obfuscated, as transformLinkname silently left
strings.IndexByte untouched. It is a problem when IndexByte does get
obfuscated. Make that kind of listPackage error visible, and fix it.
reflect.txtar started failing with "unreachable method" runtime throws.
It's not clear to me why; it appears that GOGARBLE=* makes the linker
think that ExportedMethodName is suddenly unreachable.
Work around the problem by making the method explicitly reachable,
and leave a TODO as a reminder to investigate.
Finally, gogarble.txtar no longer needs to test for GOPRIVATE.
The rest of the test is left the same, as we still want the various
values for GOGARBLE to continue to work just like before.
Fixes#594.
Since #550 in June 2022, we have stopped considering `//export`
directives when deciding what names to obfuscate.
We forgot to remove the caveat from the README.
Make the literals section easier to follow by using the word
"expression" more consistently.
The tiny section was misleading as well, as the README made it seem like
position information was stripped both with and without the flag.
Clarify what happens with that information in each case.
Now that we've released v0.6.0, that will be the last feature release to
feature support for Go 1.17. The upcoming v0.7.0 will be Go 1.18+.
Code-wise, the cleanup here isn't super noticeable,
but it will be easier to work on features like VCS-aware version
information and generics support without worrying about Go 1.17.
Plus, now CI is back to being much faster.
Note how "go 1.18" in go.mod makes "go mod tidy" more aggressive.
The default behavior of garble is to seed via the build inputs,
including the build IDs of the entire Go build of each package.
This works well as a default, and does give us determinism,
but it means that building for different platforms
will result in different obfuscation per platform.
Instead, when -seed is provided, don't use any other hash seed or salt.
This means that a particular Go name will be obfuscated the same way
as long as the seed, package path, and name itself remain constant.
In other words, when the user supplies a custom -seed,
we assume they know what they're doing in terms of storage and rotation.
Expand the README docs with more examples and detail.
Fixes#449.
The section said that "garble build" should take about twice as long as
"go build", which is typically true, with the glaring exception that is
the first "garble build" after normal use of "go build".
This scenario will be rather common for first-time users,
and the difference will be most noticeable with large projects.
A large build may take about a second to rebuild with a warm cache,
but could take tens or even hundreds of seconds to rebuild from scratch.
See #479.
Piggybacking off of GOPRIVATE is great for a number of reasons:
* People tend to obfuscate private code, whose package paths will
generally be in GOPRIVATE already
* Its meaning and syntax are well understood
* It allows all the flexibility we need without adding our own env var
or config option
However, using GOPRIVATE directly has one main drawback.
It's fairly common to also want to obfuscate public dependencies,
to make the code in private packages even harder to follow.
However, using "GOPRIVATE=*" will result in two main downsides:
* GONOPROXY defaults to GOPRIVATE, so the proxy would be entirely disabled.
Downloading modules, such as when adding or updating dependencies,
or when the local cache is cold, can be less reliable.
* GONOSUMDB defaults to GOPRIVATE, so the sumdb would be entirely disabled.
Adding entries to go.sum, such as when adding or updating dependencies,
can be less secure.
We will continue to consume GOPRIVATE as a fallback,
but we now expect users to set GOGARBLE instead.
The new logic is documented in the README.
While here, rewrite some uses of "private" with "to obfuscate",
to make the code easier to follow and harder to misunderstand.
Fixes#276.
We recently improved the automatic detection of reflection.
However, note that a hint may still be needed in some cases,
in particular when the reflection usage happens in a downstream package.
Keep making that suggestion.
Also mention that we obfuscate one package at a time.
I've also written down #406 with more ideas on how we could polish this
rough edge for end users in the future.
Fixes#403.
We can now use pruned module graphs in go.mod files,
and we no longer need to worry about runtime/internal/sys.
Note that I had to update testdata/mod slightly,
as the new pruned module graphs algorithm downloads an extra go.mod file.
This change also paves the way towards future Go 1.18 support.
Thanks to lu4p for cleaning up two TODOs as well.
Co-Authored-By: lu4p <lu4p@pm.me>
Functions which use reflection on one of their parameters are,
now added to knownReflectAPIs automatically.
This makes most explicit hints for reflection redundant.
Simple protobuf code now works correctly when obfuscated.
Fixes#162Fixes#373
Now that Go 1.17.x is out,
we no longer need to worry about users on Go 1.15.x.
Since Go 1.16, the best way to install programs has been "go install":
https://golang.org/doc/go1.16#go-command
This method does not interfere with the current module,
and allows selecting a version such as "latest" or "master".
Before, we would just notice direct calls to reflect's TypeOf and
ValueOf. Any other uses of reflection, such as encoding/json or
google.golang.org/protobuf, would require hints as documented by the
README.
Issue #162 outlines some ways we could fix this issue in a general way,
automatically detecting what functions use reflection on their parameters,
even for third party API funcs.
However, that goal is pretty significant in terms of code and effort.
As a temporary improvement, we can expand the list of "known" reflection
APIs via a static table.
Since this table is keyed by "func full name" strings, we could
potentially include third party APIs, such as:
google.golang.org/protobuf/proto.Marshal
However, for now simply include all the std APIs we know about.
If we fail to do the proper fix for automatic detection in the future,
we can then fall back to expanding this global table for third parties.
Update the README's docs, to clarify that the hint is not always
necessary anymore.
Also update the reflect.txt test to stop using the hint for encoding/json,
and to also start testing text/template with a method call.
While at it, I noticed that we weren't testing the println outputs,
as they'd go to stderr - fix that too.
Updates #162.
Otherwise, the added test case would fail, as we don't modify the C code
and so there would be a name mismatch.
In the far future we might start modifying Go names in C code,
similar to what we did for Go assembly,
but right now that seems out of scope and too complex.
An easier fix is to simply record those (hopefully few) names in ignoreObjects.
While at it, recordReflectArgs started to really outgrow its name, as it
also collected expressions used as constants for literal obfuscation.
Give it a better name.
Fixes#366.
The sections are a bit more organized now. We add better docs for flags,
a new section on build speed, and we also remove the "withgo" helper as
it seems a bit over the top.
Fixes#275.
Historically, it was impossible to rename those funcs as the
implementation was in assembly files, and we only transformed Go code.
Now that transformAsm exists, it's only about fifty lines to do some
very basic parsing and rewriting of assembly files.
This fixes the obfuscated builds of multiple std packages, including a
few dependencies of net/http, since they included assembly funcs which
called pure Go functions. Those pure Go functions had their names
obfuscated, breaking the call sites in assembly.
Fixes#258.
Fixes#261.
This mainly cleans up the few bits of code where we explicitly kept
support for Go 1.15.x. With v0.1.0 released, we can drop support now,
since the next v0.2.0 release will only support Go 1.16.x.
Also updates all modules, including test ones, to 'go 1.16'.
Note that the TOOLEXEC_IMPORTPATH refactor is not done here, despite all
the TODOs about doing so when we drop 1.15 support. This is because that
refactor needs to be done carefully and might have side effects, so it's
best to keep it to a separate commit.
Finally, update the deps.
First, write a changelog file. We will use GitHub releases, but the
content in those is not stored in git nor is it portable or machine
readable. The canonical place for the changelog is here.
Second, disable 'garble test', as it is entirely broken. Issue #241
tracks fixing and re-enabling it, which will most likely happen for the
next release.
Third, disable the undocumented 'garble list'. This was added as part of
'garble reverse', but it never got used. I can't think of any reason why
any end user would prefer it over 'go list', either.
'garble reverse' remains enabled, but undocumented as it isn't fully
functional yet. Until it supports position information, it's not
particularly useful to end users. But it's not broken either, so it can
remain where it is.
Fourth, update the '-tiny' size reduction numbers in the README. Since
we removed the in-place modification of object files, we are no longer
able to do such an aggressive stripping of info. Garble itself drops in
size by 2%, so replace the old 6-10% estimate by 2-5%. We probably will
gain some of this back in the near future.
Finally, fix the indentation formatting of the README to consistently
use tabs.
This was pretty much just fixing the README and closing the issue. The
only other noteworthy user-facing change is that, if the Go version is
detected to be too old, we now suggest 1.16.x instead of 1.15.x.
While at it, refactor goversion.txt a bit. I wanted it to print a
clearer "mocking the go build" error if another command was used like
"go build", but I didn't want to learn BAT. So, instead use a simple Go
program and build it, which will work on all platforms. The added
"go build" step barely takes 100ms on my machine, given how simple the
program is.
The [short] line also doesn't seem necessary to me. The entire script
runs in under 200ms for me, so it's well within the realm of "short", at
least compared to many of the other test scripts.
Fixes#124.
This is a bit tricky right now, but will hopefully become simpler if
https://github.com/golang/go/issues/44329 gets approved and implemented.
Until then, suggest ways to do it in POSIX Shell.
While at it, stop using "garble" as a verb, and just use "obfuscate".
It's a bit silly if we say "garble garbles code"; it's much more
intuitive and less repetitive to say "garble obfuscates code".
The code still says "garble X" and "X was garbled" in a few places, and
we should fix those, but at least they are not publicly visible.
Since that version is out now, this will help avoid confusion with
newcomers. The issue is pinned now, too.
While at it, stop saying why 1.14.x is no longer supported, as it's not
a maintained version of Go at this point anymore.
For #124.
In 90fa325da7, the obfuscation logic was changed to use hashes for
exported names, but incremental names starting at just one letter for
unexported names. Presumably, this was done for the sake of binary size.
I argue that this is not a good idea for the default mode for a number
of reasons:
1) It makes reversing of stack traces nearly impossible for unexported
names, since replacing an obfuscated name "c" with "originalName"
would trigger too many false positives by matching single characters.
2) Exported and unexported names aren't different. We need to know how
names were obfuscated at a later time in both cases, thanks to use
cases like -ldflags=-X. Using short names for one but not the other
doesn't make a lot of sense, and makes the logic inconsistent.
3) Shaving off three bytes for unexported names doesn't seem like a huge
deal for the default mode, when we already have -tiny to optimize for
size.
This saves us a bit of work, but most importantly, simplifies the
obfuscation state as we no longer need to carry privateNameMap between
the compile and link stages.
name old time/op new time/op delta
Build-8 153ms ± 2% 150ms ± 2% ~ (p=0.065 n=6+6)
name old bin-B new bin-B delta
Build-8 7.09M ± 0% 7.08M ± 0% -0.24% (p=0.002 n=6+6)
name old sys-time/op new sys-time/op delta
Build-8 296ms ± 5% 277ms ± 6% -6.50% (p=0.026 n=6+6)
name old user-time/op new user-time/op delta
Build-8 562ms ± 1% 558ms ± 3% ~ (p=0.329 n=5+6)
Note that I do not oppose using short names for both exported and
unexported names in the future for -tiny, since reversing of stack
traces will by design not work there. The code can be resurrected from
the git history if we want to improve -tiny that way in the future, as
we'd need to store state in header files again.
Another major cleanup we can do here is to no longer use the
garbledImports map. From a look at obfuscateImports, we hash a package's
import path with its action ID, much like exported names, so we can
simply re-do that hashing for the linker's -X flag.
garbledImports does have some logic to handle duplicate package names,
but it's worth noting that should not affect package paths, as they are
always unique. That area of code could probably do with some
simplification in the future, too.
While at it, make hashWith panic if either parameter is empty.
obfuscateImports was hashing the main package path without a salt due to
a bug, so we want to catch those in the future.
Finally, make some tiny spacing and typo tweaks to the README.
As per the discussion in https://github.com/golang/go/issues/41145, it
turns out that we don't need special support for build caching in
-toolexec. We can simply modify the behavior of "[...]/compile -V=full"
and "[...]/link -V=full" so that they include garble's own version and
options in the printed build ID.
The part of the build ID that matters is the last, since it's the
"content ID" which is used to work out whether there is a need to redo
the action (build) or not. Since cmd/go parses the last word in the
output as "buildID=...", we simply add "+garble buildID=_/_/_/${hash}".
The slashes let us imitate a full binary build ID, but we assume that
the other components such as the action ID are not necessary, since the
only reader here is cmd/go and it only consumes the content ID.
The reported content ID includes the tool's original content ID,
garble's own content ID from the built binary, and the garble options
which modify how we obfuscate code. If any of the three changes, we
should use a different build cache key. GOPRIVATE also affects caching,
since a different GOPRIVATE value means that we might have to garble a
different set of packages.
Include tests, which mainly check that 'garble build -v' prints package
lines when we expect to always need to rebuild packages, and that it
prints nothing when we should be reusing the build cache even when the
built binary is missing.
After this change, 'go test' on Go 1.15.2 stabilizes at about 8s on my
machine, whereas it used to be at around 25s before.
Rework the features section in the README, leaving optional features at
the end of the list. Simplify the caveats list, too; the build cache and
exported field/method bits only need one point each. Overall, the
section was far too wordy for little reason.
Also redo the help text a bit. There's now a line to briefly introduce
the tool, as well as a link to the README with all the details. Finally,
the flags have shorter and more consistent help strings.
While at it, remove two unused global vars as spotted by staticcheck.
Finally, finally this is done. This allows import paths to be obfuscated by modifying
object/archive files and garbling import paths contained within. The bulk of the
code that makes parsing and writing Go object/archive files possible lives at
https://github.com/Binject/debug/tree/master/goobj2, which I wrote as well.
I have tested by garbling and checking for import paths via strings and grep
(in order of difficulty) https://github.com/lu4p/binclude, garble itself, and
https://github.com/dominikh/go-tools/tree/master/cmd/staticcheck.
This only supports object/archive files produced from the Go 1.15 compiler.
The object file format changed at 1.15, and 1.14 and earlier is not supported.
Fixes#13.
Most notably, x/mod now includes the GOPRIVATE pattern-matching API we
were copying before, so we can use it directly.
Also bump the Go version requirement to 1.15, in preparation for the
import path obfuscation PR, and don't let the gotip job fail the entire
workflow.
Daniel Martí
pagran
lu4p
Damien Lespiau
Andrew LeFevre