gecko-dev/dom/security/SRICheck.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

511 строки
18 KiB
C++
Исходник Обычный вид История

/* -*- Mode: C++; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* vim: set ts=8 sts=2 et sw=2 tw=80: */
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
#include "SRICheck.h"
#include "mozilla/Base64.h"
#include "mozilla/LoadTainting.h"
#include "mozilla/Logging.h"
#include "mozilla/Preferences.h"
#include "mozilla/dom/SRILogHelper.h"
#include "mozilla/dom/SRIMetadata.h"
#include "nsContentUtils.h"
#include "nsIChannel.h"
#include "nsIConsoleReportCollector.h"
#include "nsIProtocolHandler.h"
#include "nsIScriptError.h"
#include "nsIIncrementalStreamLoader.h"
#include "nsIURI.h"
#include "nsNetUtil.h"
#include "nsWhitespaceTokenizer.h"
#define SRIVERBOSE(args) \
MOZ_LOG(SRILogHelper::GetSriLog(), mozilla::LogLevel::Verbose, args)
#define SRILOG(args) \
MOZ_LOG(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug, args)
#define SRIERROR(args) \
MOZ_LOG(SRILogHelper::GetSriLog(), mozilla::LogLevel::Error, args)
namespace mozilla {
namespace dom {
/**
* Returns whether or not the sub-resource about to be loaded is eligible
* for integrity checks. If it's not, the checks will be skipped and the
* sub-resource will be loaded.
*/
static nsresult IsEligible(nsIChannel* aChannel,
mozilla::LoadTainting aTainting,
const nsACString& aSourceFileURI,
nsIConsoleReportCollector* aReporter) {
NS_ENSURE_ARG_POINTER(aReporter);
if (!aChannel) {
SRILOG(("SRICheck::IsEligible, null channel"));
return NS_ERROR_SRI_NOT_ELIGIBLE;
}
// Was the sub-resource loaded via CORS?
if (aTainting == LoadTainting::CORS) {
SRILOG(("SRICheck::IsEligible, CORS mode"));
return NS_OK;
}
nsCOMPtr<nsIURI> finalURI;
nsresult rv = NS_GetFinalChannelURI(aChannel, getter_AddRefs(finalURI));
NS_ENSURE_SUCCESS(rv, rv);
nsCOMPtr<nsIURI> originalURI;
rv = aChannel->GetOriginalURI(getter_AddRefs(originalURI));
NS_ENSURE_SUCCESS(rv, rv);
nsAutoCString requestSpec;
rv = originalURI->GetSpec(requestSpec);
NS_ENSURE_SUCCESS(rv, rv);
if (MOZ_LOG_TEST(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug)) {
SRILOG(("SRICheck::IsEligible, requestURI=%s; finalURI=%s",
requestSpec.get(),
finalURI ? finalURI->GetSpecOrDefault().get() : ""));
}
// Is the sub-resource same-origin?
if (aTainting == LoadTainting::Basic) {
SRILOG(("SRICheck::IsEligible, same-origin"));
return NS_OK;
}
SRILOG(("SRICheck::IsEligible, NOT same origin"));
NS_ConvertUTF8toUTF16 requestSpecUTF16(requestSpec);
nsTArray<nsString> params;
params.AppendElement(requestSpecUTF16);
aReporter->AddConsoleReport(
nsIScriptError::errorFlag, NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("IneligibleResource"),
const_cast<const nsTArray<nsString>&>(params));
return NS_ERROR_SRI_NOT_ELIGIBLE;
}
/* static */ nsresult SRICheck::IntegrityMetadata(
const nsAString& aMetadataList, const nsACString& aSourceFileURI,
nsIConsoleReportCollector* aReporter, SRIMetadata* outMetadata) {
NS_ENSURE_ARG_POINTER(outMetadata);
NS_ENSURE_ARG_POINTER(aReporter);
MOZ_ASSERT(outMetadata->IsEmpty()); // caller must pass empty metadata
if (!Preferences::GetBool("security.sri.enable", false)) {
SRILOG(("SRICheck::IntegrityMetadata, sri is disabled (pref)"));
return NS_ERROR_SRI_DISABLED;
}
// put a reasonable bound on the length of the metadata
Bug 1402247 - Use encoding_rs for XPCOM string encoding conversions. r=Nika,erahm,froydnj. Correctness improvements: * UTF errors are handled safely per spec instead of dangerously truncating strings. * There are fewer converter implementations. Performance improvements: * The old code did exact buffer length math, which meant doing UTF math twice on each input string (once for length calculation and another time for conversion). Exact length math is more complicated when handling errors properly, which the old code didn't do. The new code does UTF math on the string content only once (when converting) but risks allocating more than once. There are heuristics in place to lower the probability of reallocation in cases where the double math avoidance isn't enough of a saving to absorb an allocation and memcpy. * Previously, in UTF-16 <-> UTF-8 conversions, an ASCII prefix was optimized but a single non-ASCII code point pessimized the rest of the string. The new code tries to get back on the fast ASCII path. * UTF-16 to Latin1 conversion guarantees less about handling of out-of-range input to eliminate an operation from the inner loop on x86/x86_64. * When assigning to a pre-existing string, the new code tries to reuse the old buffer instead of first releasing the old buffer and then allocating a new one. * When reallocating from the new code, the memcpy covers only the data that is part of the logical length of the old string instead of memcpying the whole capacity. (For old callers old excess memcpy behavior is preserved due to bogus callers. See bug 1472113.) * UTF-8 strings in XPConnect that are in the Latin1 range are passed to SpiderMonkey as Latin1. New features: * Conversion between UTF-8 and Latin1 is added in order to enable faster future interop between Rust code (or otherwise UTF-8-using code) and text node and SpiderMonkey code that uses Latin1. MozReview-Commit-ID: JaJuExfILM9
2018-07-06 10:44:43 +03:00
NS_ConvertUTF16toUTF8 metadataList(aMetadataList);
if (metadataList.Length() > SRICheck::MAX_METADATA_LENGTH) {
metadataList.Truncate(SRICheck::MAX_METADATA_LENGTH);
}
SRILOG(("SRICheck::IntegrityMetadata, metadataList=%s", metadataList.get()));
// the integrity attribute is a list of whitespace-separated hashes
// and options so we need to look at them one by one and pick the
// strongest (valid) one
nsCWhitespaceTokenizer tokenizer(metadataList);
nsAutoCString token;
for (uint32_t i = 0;
tokenizer.hasMoreTokens() && i < SRICheck::MAX_METADATA_TOKENS; ++i) {
token = tokenizer.nextToken();
SRIMetadata metadata(token);
if (metadata.IsMalformed()) {
NS_ConvertUTF8toUTF16 tokenUTF16(token);
nsTArray<nsString> params;
params.AppendElement(tokenUTF16);
aReporter->AddConsoleReport(
nsIScriptError::warningFlag,
NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("MalformedIntegrityHash"),
const_cast<const nsTArray<nsString>&>(params));
} else if (!metadata.IsAlgorithmSupported()) {
nsAutoCString alg;
metadata.GetAlgorithm(&alg);
NS_ConvertUTF8toUTF16 algUTF16(alg);
nsTArray<nsString> params;
params.AppendElement(algUTF16);
aReporter->AddConsoleReport(
nsIScriptError::warningFlag,
NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("UnsupportedHashAlg"),
const_cast<const nsTArray<nsString>&>(params));
}
nsAutoCString alg1, alg2;
if (MOZ_LOG_TEST(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug)) {
outMetadata->GetAlgorithm(&alg1);
metadata.GetAlgorithm(&alg2);
}
if (*outMetadata == metadata) {
SRILOG(("SRICheck::IntegrityMetadata, alg '%s' is the same as '%s'",
alg1.get(), alg2.get()));
*outMetadata += metadata; // add new hash to strongest metadata
} else if (*outMetadata < metadata) {
SRILOG(("SRICheck::IntegrityMetadata, alg '%s' is weaker than '%s'",
alg1.get(), alg2.get()));
*outMetadata = metadata; // replace strongest metadata with current
}
}
outMetadata->mIntegrityString = aMetadataList;
if (MOZ_LOG_TEST(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug)) {
if (outMetadata->IsValid()) {
nsAutoCString alg;
outMetadata->GetAlgorithm(&alg);
SRILOG(("SRICheck::IntegrityMetadata, using a '%s' hash", alg.get()));
} else if (outMetadata->IsEmpty()) {
SRILOG(("SRICheck::IntegrityMetadata, no metadata"));
} else {
SRILOG(("SRICheck::IntegrityMetadata, no valid metadata found"));
}
}
return NS_OK;
}
//////////////////////////////////////////////////////////////
//
//////////////////////////////////////////////////////////////
SRICheckDataVerifier::SRICheckDataVerifier(const SRIMetadata& aMetadata,
const nsACString& aSourceFileURI,
nsIConsoleReportCollector* aReporter)
: mCryptoHash(nullptr),
mBytesHashed(0),
mHashLength(0),
mHashType('\0'),
mInvalidMetadata(false),
mComplete(false) {
MOZ_ASSERT(!aMetadata.IsEmpty()); // should be checked by caller
// IntegrityMetadata() checks this and returns "no metadata" if
// it's disabled so we should never make it this far
MOZ_ASSERT(Preferences::GetBool("security.sri.enable", false));
MOZ_ASSERT(aReporter);
if (!aMetadata.IsValid()) {
nsTArray<nsString> params;
aReporter->AddConsoleReport(nsIScriptError::warningFlag,
NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES,
aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("NoValidMetadata"),
const_cast<const nsTArray<nsString>&>(params));
mInvalidMetadata = true;
return; // ignore invalid metadata for forward-compatibility
}
aMetadata.GetHashType(&mHashType, &mHashLength);
}
nsresult SRICheckDataVerifier::EnsureCryptoHash() {
MOZ_ASSERT(!mInvalidMetadata);
if (mCryptoHash) {
return NS_OK;
}
nsresult rv;
nsCOMPtr<nsICryptoHash> cryptoHash =
do_CreateInstance("@mozilla.org/security/hash;1", &rv);
NS_ENSURE_SUCCESS(rv, rv);
rv = cryptoHash->Init(mHashType);
NS_ENSURE_SUCCESS(rv, rv);
mCryptoHash = cryptoHash;
return NS_OK;
}
nsresult SRICheckDataVerifier::Update(uint32_t aStringLen,
const uint8_t* aString) {
NS_ENSURE_ARG_POINTER(aString);
if (mInvalidMetadata) {
return NS_OK; // ignoring any data updates, see mInvalidMetadata usage
}
nsresult rv;
rv = EnsureCryptoHash();
NS_ENSURE_SUCCESS(rv, rv);
mBytesHashed += aStringLen;
return mCryptoHash->Update(aString, aStringLen);
}
nsresult SRICheckDataVerifier::Finish() {
if (mInvalidMetadata || mComplete) {
return NS_OK; // already finished or invalid metadata
}
nsresult rv;
rv = EnsureCryptoHash(); // we need computed hash even for 0-length data
NS_ENSURE_SUCCESS(rv, rv);
rv = mCryptoHash->Finish(false, mComputedHash);
mCryptoHash = nullptr;
mComplete = true;
return rv;
}
nsresult SRICheckDataVerifier::VerifyHash(
const SRIMetadata& aMetadata, uint32_t aHashIndex,
const nsACString& aSourceFileURI, nsIConsoleReportCollector* aReporter) {
NS_ENSURE_ARG_POINTER(aReporter);
nsAutoCString base64Hash;
aMetadata.GetHash(aHashIndex, &base64Hash);
SRILOG(("SRICheckDataVerifier::VerifyHash, hash[%u]=%s", aHashIndex,
base64Hash.get()));
nsAutoCString binaryHash;
if (NS_WARN_IF(NS_FAILED(Base64Decode(base64Hash, binaryHash)))) {
nsTArray<nsString> params;
aReporter->AddConsoleReport(
nsIScriptError::errorFlag, NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("InvalidIntegrityBase64"),
const_cast<const nsTArray<nsString>&>(params));
return NS_ERROR_SRI_CORRUPT;
}
uint32_t hashLength;
int8_t hashType;
aMetadata.GetHashType(&hashType, &hashLength);
if (binaryHash.Length() != hashLength) {
nsTArray<nsString> params;
aReporter->AddConsoleReport(
nsIScriptError::errorFlag, NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("InvalidIntegrityLength"),
const_cast<const nsTArray<nsString>&>(params));
return NS_ERROR_SRI_CORRUPT;
}
if (MOZ_LOG_TEST(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug)) {
nsAutoCString encodedHash;
nsresult rv = Base64Encode(mComputedHash, encodedHash);
if (NS_SUCCEEDED(rv)) {
SRILOG(("SRICheckDataVerifier::VerifyHash, mComputedHash=%s",
encodedHash.get()));
}
}
if (!binaryHash.Equals(mComputedHash)) {
SRILOG(("SRICheckDataVerifier::VerifyHash, hash[%u] did not match",
aHashIndex));
return NS_ERROR_SRI_CORRUPT;
}
SRILOG(("SRICheckDataVerifier::VerifyHash, hash[%u] verified successfully",
aHashIndex));
return NS_OK;
}
nsresult SRICheckDataVerifier::Verify(const SRIMetadata& aMetadata,
nsIChannel* aChannel,
const nsACString& aSourceFileURI,
nsIConsoleReportCollector* aReporter) {
NS_ENSURE_ARG_POINTER(aReporter);
if (MOZ_LOG_TEST(SRILogHelper::GetSriLog(), mozilla::LogLevel::Debug)) {
nsAutoCString requestURL;
nsCOMPtr<nsIRequest> request = aChannel;
request->GetName(requestURL);
SRILOG(("SRICheckDataVerifier::Verify, url=%s (length=%zu)",
requestURL.get(), mBytesHashed));
}
nsresult rv = Finish();
NS_ENSURE_SUCCESS(rv, rv);
nsCOMPtr<nsILoadInfo> loadInfo = aChannel->GetLoadInfo();
NS_ENSURE_TRUE(loadInfo, NS_ERROR_FAILURE);
LoadTainting tainting = loadInfo->GetTainting();
if (NS_FAILED(IsEligible(aChannel, tainting, aSourceFileURI, aReporter))) {
return NS_ERROR_SRI_NOT_ELIGIBLE;
}
if (mInvalidMetadata) {
return NS_OK; // ignore invalid metadata for forward-compatibility
}
for (uint32_t i = 0; i < aMetadata.HashCount(); i++) {
if (NS_SUCCEEDED(VerifyHash(aMetadata, i, aSourceFileURI, aReporter))) {
return NS_OK; // stop at the first valid hash
}
}
nsAutoCString alg;
aMetadata.GetAlgorithm(&alg);
NS_ConvertUTF8toUTF16 algUTF16(alg);
nsTArray<nsString> params;
params.AppendElement(algUTF16);
aReporter->AddConsoleReport(
nsIScriptError::errorFlag, NS_LITERAL_CSTRING("Sub-resource Integrity"),
nsContentUtils::eSECURITY_PROPERTIES, aSourceFileURI, 0, 0,
NS_LITERAL_CSTRING("IntegrityMismatch"),
const_cast<const nsTArray<nsString>&>(params));
return NS_ERROR_SRI_CORRUPT;
}
uint32_t SRICheckDataVerifier::DataSummaryLength() {
MOZ_ASSERT(!mInvalidMetadata);
return sizeof(mHashType) + sizeof(mHashLength) + mHashLength;
}
uint32_t SRICheckDataVerifier::EmptyDataSummaryLength() {
return sizeof(int8_t) + sizeof(uint32_t);
}
nsresult SRICheckDataVerifier::DataSummaryLength(uint32_t aDataLen,
const uint8_t* aData,
uint32_t* length) {
*length = 0;
NS_ENSURE_ARG_POINTER(aData);
// we expect to always encode an SRI, even if it is empty or incomplete
if (aDataLen < EmptyDataSummaryLength()) {
SRILOG(
("SRICheckDataVerifier::DataSummaryLength, encoded length[%u] is too "
"small",
aDataLen));
return NS_ERROR_SRI_IMPORT;
}
// decode the content of the buffer
size_t offset = sizeof(mHashType);
decltype(mHashLength) len = 0;
memcpy(&len, &aData[offset], sizeof(mHashLength));
offset += sizeof(mHashLength);
SRIVERBOSE(
("SRICheckDataVerifier::DataSummaryLength, header {%x, %x, %x, %x, %x, "
"...}",
aData[0], aData[1], aData[2], aData[3], aData[4]));
if (offset + len > aDataLen) {
SRILOG(
("SRICheckDataVerifier::DataSummaryLength, encoded length[%u] overflow "
"the buffer size",
aDataLen));
SRIVERBOSE(("SRICheckDataVerifier::DataSummaryLength, offset[%u], len[%u]",
uint32_t(offset), uint32_t(len)));
return NS_ERROR_SRI_IMPORT;
}
*length = uint32_t(offset + len);
return NS_OK;
}
nsresult SRICheckDataVerifier::ImportDataSummary(uint32_t aDataLen,
const uint8_t* aData) {
MOZ_ASSERT(!mInvalidMetadata); // mHashType and mHashLength should be valid
MOZ_ASSERT(!mCryptoHash); // EnsureCryptoHash should not have been called
NS_ENSURE_ARG_POINTER(aData);
if (mInvalidMetadata) {
return NS_OK; // ignoring any data updates, see mInvalidMetadata usage
}
// we expect to always encode an SRI, even if it is empty or incomplete
if (aDataLen < DataSummaryLength()) {
SRILOG(
("SRICheckDataVerifier::ImportDataSummary, encoded length[%u] is too "
"small",
aDataLen));
return NS_ERROR_SRI_IMPORT;
}
SRIVERBOSE(
("SRICheckDataVerifier::ImportDataSummary, header {%x, %x, %x, %x, %x, "
"...}",
aData[0], aData[1], aData[2], aData[3], aData[4]));
// decode the content of the buffer
size_t offset = 0;
decltype(mHashType) hashType;
memcpy(&hashType, &aData[offset], sizeof(mHashType));
if (hashType != mHashType) {
SRILOG(
("SRICheckDataVerifier::ImportDataSummary, hash type[%d] does not "
"match[%d]",
hashType, mHashType));
return NS_ERROR_SRI_UNEXPECTED_HASH_TYPE;
}
offset += sizeof(mHashType);
decltype(mHashLength) hashLength;
memcpy(&hashLength, &aData[offset], sizeof(mHashLength));
if (hashLength != mHashLength) {
SRILOG(
("SRICheckDataVerifier::ImportDataSummary, hash length[%d] does not "
"match[%d]",
hashLength, mHashLength));
return NS_ERROR_SRI_UNEXPECTED_HASH_TYPE;
}
offset += sizeof(mHashLength);
// copy the hash to mComputedHash, as-if we had finished streaming the bytes
mComputedHash.Assign(reinterpret_cast<const char*>(&aData[offset]),
mHashLength);
mCryptoHash = nullptr;
mComplete = true;
return NS_OK;
}
nsresult SRICheckDataVerifier::ExportDataSummary(uint32_t aDataLen,
uint8_t* aData) {
MOZ_ASSERT(!mInvalidMetadata); // mHashType and mHashLength should be valid
MOZ_ASSERT(mComplete); // finished streaming
NS_ENSURE_ARG_POINTER(aData);
NS_ENSURE_TRUE(aDataLen >= DataSummaryLength(), NS_ERROR_INVALID_ARG);
// serialize the hash in the buffer
size_t offset = 0;
memcpy(&aData[offset], &mHashType, sizeof(mHashType));
offset += sizeof(mHashType);
memcpy(&aData[offset], &mHashLength, sizeof(mHashLength));
offset += sizeof(mHashLength);
SRIVERBOSE(
("SRICheckDataVerifier::ExportDataSummary, header {%x, %x, %x, %x, %x, "
"...}",
aData[0], aData[1], aData[2], aData[3], aData[4]));
// copy the hash to mComputedHash, as-if we had finished streaming the bytes
nsCharTraits<char>::copy(reinterpret_cast<char*>(&aData[offset]),
mComputedHash.get(), mHashLength);
return NS_OK;
}
nsresult SRICheckDataVerifier::ExportEmptyDataSummary(uint32_t aDataLen,
uint8_t* aData) {
NS_ENSURE_ARG_POINTER(aData);
NS_ENSURE_TRUE(aDataLen >= EmptyDataSummaryLength(), NS_ERROR_INVALID_ARG);
// serialize an unknown hash in the buffer, to be able to skip it later
size_t offset = 0;
memset(&aData[offset], 0, sizeof(mHashType));
offset += sizeof(mHashType);
memset(&aData[offset], 0, sizeof(mHashLength));
offset += sizeof(mHashLength);
SRIVERBOSE(
("SRICheckDataVerifier::ExportEmptyDataSummary, header {%x, %x, %x, %x, "
"%x, ...}",
aData[0], aData[1], aData[2], aData[3], aData[4]));
return NS_OK;
}
} // namespace dom
} // namespace mozilla