crashpad/util/posix/scoped_mmap_test.cc

409 lines
14 KiB
C++
Raw Normal View History

// Copyright 2017 The Crashpad Authors. All rights reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
#include "util/posix/scoped_mmap.h"
#include <stdint.h>
#include <sys/types.h>
#include <unistd.h>
#include "base/numerics/safe_conversions.h"
#include "base/rand_util.h"
#include "base/stl_util.h"
#include "base/strings/stringprintf.h"
#include "gtest/gtest.h"
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
#include "test/gtest_death.h"
namespace crashpad {
namespace test {
namespace {
bool ScopedMmapResetMmap(ScopedMmap* mapping, size_t len) {
return mapping->ResetMmap(
nullptr, len, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0);
}
void* BareMmap(size_t len) {
return mmap(
nullptr, len, PROT_READ | PROT_WRITE, MAP_PRIVATE | MAP_ANON, -1, 0);
}
// A weird class. This is used to test that memory-mapped regions are freed
// as expected by calling munmap(). This is difficult to test well because once
// a region has been unmapped, the address space it formerly occupied becomes
// eligible for reuse.
//
// The strategy taken here is that a random 64-bit cookie value is written into
// a mapped region by SetUp(). While the mapping is active, Check() should not
// crash, or for a gtest expectation, Expected() and Observed() should not crash
// and should be equal. After the region is unmapped, Check() should crash,
// either because the region has been unmapped and the address not reused, the
// address has been reused but is protected against reading (unlikely), or
// because the address has been reused but the cookie value is no longer present
// there.
class TestCookie {
public:
// A weird constructor for a weird class. The member variable initialization
// assures that Check() wont crash if called on an object that hasnt had
// SetUp() called on it.
explicit TestCookie() : address_(&cookie_), cookie_(0) {}
~TestCookie() {}
void SetUp(uint64_t* address) {
address_ = address, cookie_ = base::RandUint64();
*address_ = cookie_;
}
uint64_t Expected() const { return cookie_; }
uint64_t Observed() const { return *address_; }
void Check() const {
if (Observed() != Expected()) {
__builtin_trap();
}
}
private:
uint64_t* address_;
uint64_t cookie_;
DISALLOW_COPY_AND_ASSIGN(TestCookie);
};
TEST(ScopedMmap, Mmap) {
TestCookie cookie;
ScopedMmap mapping;
EXPECT_FALSE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 0u);
ASSERT_TRUE(mapping.Reset());
EXPECT_FALSE(mapping.is_valid());
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kPageSize);
cookie.SetUp(mapping.addr_as<uint64_t*>());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(cookie.Observed(), cookie.Expected());
ASSERT_TRUE(mapping.Reset());
EXPECT_FALSE(mapping.is_valid());
}
TEST(ScopedMmapDeathTest, Destructor) {
TestCookie cookie;
{
ScopedMmap mapping;
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kPageSize);
cookie.SetUp(mapping.addr_as<uint64_t*>());
}
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookie.Check(), "");
}
TEST(ScopedMmapDeathTest, Reset) {
ScopedMmap mapping;
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kPageSize);
TestCookie cookie;
cookie.SetUp(mapping.addr_as<uint64_t*>());
ASSERT_TRUE(mapping.Reset());
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookie.Check(), "");
}
TEST(ScopedMmapDeathTest, ResetAddrLen_Shrink) {
ScopedMmap mapping;
// Start with three pages mapped.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, 3 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 3 * kPageSize);
TestCookie cookies[3];
for (size_t index = 0; index < base::size(cookies); ++index) {
cookies[index].SetUp(reinterpret_cast<uint64_t*>(
mapping.addr_as<uintptr_t>() + index * kPageSize));
}
// Reset to the second page. The first and third pages should be unmapped.
void* const new_addr =
reinterpret_cast<void*>(mapping.addr_as<uintptr_t>() + kPageSize);
ASSERT_TRUE(mapping.ResetAddrLen(new_addr, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), new_addr);
EXPECT_EQ(mapping.len(), kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(cookies[1].Observed(), cookies[1].Expected());
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookies[0].Check(), "");
EXPECT_DEATH_CRASH(cookies[2].Check(), "");
}
TEST(ScopedMmap, ResetAddrLen_Grow) {
// Start with three pages mapped, but ScopedMmap only aware of the the second
// page.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
void* pages = BareMmap(3 * kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
ASSERT_NE(pages, MAP_FAILED);
ScopedMmap mapping;
void* const old_addr =
reinterpret_cast<void*>(reinterpret_cast<uintptr_t>(pages) + kPageSize);
ASSERT_TRUE(mapping.ResetAddrLen(old_addr, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), old_addr);
EXPECT_EQ(mapping.len(), kPageSize);
TestCookie cookies[3];
for (size_t index = 0; index < base::size(cookies); ++index) {
cookies[index].SetUp(reinterpret_cast<uint64_t*>(
reinterpret_cast<uintptr_t>(pages) + index * kPageSize));
}
// Reset to all three pages. Nothing should be unmapped until destruction.
ASSERT_TRUE(mapping.ResetAddrLen(pages, 3 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), pages);
EXPECT_EQ(mapping.len(), 3 * kPageSize);
for (size_t index = 0; index < base::size(cookies); ++index) {
SCOPED_TRACE(base::StringPrintf("index %zu", index));
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(cookies[index].Observed(), cookies[index].Expected());
}
}
TEST(ScopedMmapDeathTest, ResetAddrLen_MoveDownAndGrow) {
// Start with three pages mapped, but ScopedMmap only aware of the third page.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
void* pages = BareMmap(3 * kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
ASSERT_NE(pages, MAP_FAILED);
ScopedMmap mapping;
void* const old_addr = reinterpret_cast<void*>(
reinterpret_cast<uintptr_t>(pages) + 2 * kPageSize);
ASSERT_TRUE(mapping.ResetAddrLen(old_addr, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), old_addr);
EXPECT_EQ(mapping.len(), kPageSize);
TestCookie cookies[3];
for (size_t index = 0; index < base::size(cookies); ++index) {
cookies[index].SetUp(reinterpret_cast<uint64_t*>(
reinterpret_cast<uintptr_t>(pages) + index * kPageSize));
}
// Reset to the first two pages. The third page should be unmapped.
ASSERT_TRUE(mapping.ResetAddrLen(pages, 2 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), pages);
EXPECT_EQ(mapping.len(), 2 * kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(cookies[0].Observed(), cookies[0].Expected());
EXPECT_EQ(cookies[1].Observed(), cookies[1].Expected());
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookies[2].Check(), "");
}
TEST(ScopedMmapDeathTest, ResetAddrLen_MoveUpAndShrink) {
// Start with three pages mapped, but ScopedMmap only aware of the first two
// pages.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
void* pages = BareMmap(3 * kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
ASSERT_NE(pages, MAP_FAILED);
ScopedMmap mapping;
ASSERT_TRUE(mapping.ResetAddrLen(pages, 2 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), pages);
EXPECT_EQ(mapping.len(), 2 * kPageSize);
TestCookie cookies[3];
for (size_t index = 0; index < base::size(cookies); ++index) {
cookies[index].SetUp(reinterpret_cast<uint64_t*>(
reinterpret_cast<uintptr_t>(pages) + index * kPageSize));
}
// Reset to the third page. The first two pages should be unmapped.
void* const new_addr =
reinterpret_cast<void*>(mapping.addr_as<uintptr_t>() + 2 * kPageSize);
ASSERT_TRUE(mapping.ResetAddrLen(new_addr, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(mapping.addr(), new_addr);
EXPECT_EQ(mapping.len(), kPageSize);
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(cookies[2].Observed(), cookies[2].Expected());
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookies[0].Check(), "");
EXPECT_DEATH_CRASH(cookies[1].Check(), "");
}
TEST(ScopedMmapDeathTest, ResetMmap) {
ScopedMmap mapping;
// Calling ScopedMmap::ResetMmap() frees the existing mapping before
// establishing the new one, so the new one may wind up at the same address as
// the old. In fact, this is likely. Create a two-page mapping and replace it
// with a single-page mapping, so that the test can assure that the second
// page isnt mapped after establishing the second mapping.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, 2 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 2 * kPageSize);
TestCookie cookie;
cookie.SetUp(
reinterpret_cast<uint64_t*>(mapping.addr_as<char*>() + kPageSize));
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kPageSize);
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(cookie.Check(), "");
}
TEST(ScopedMmapDeathTest, NotIntegralNumberOfPages) {
ScopedMmap mapping;
EXPECT_FALSE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 0u);
ASSERT_TRUE(mapping.Reset());
EXPECT_FALSE(mapping.is_valid());
// Establishing a half-page mapping actually establishes a single page.
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
const size_t kHalfPageSize = kPageSize / 2;
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kHalfPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kHalfPageSize);
TestCookie cookie;
cookie.SetUp(mapping.addr_as<uint64_t*>());
// Shrinking a one-page mapping to a half page is a no-op.
void* orig_addr = mapping.addr();
ASSERT_TRUE(mapping.ResetAddrLen(orig_addr, kHalfPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), orig_addr);
EXPECT_EQ(mapping.len(), kHalfPageSize);
EXPECT_EQ(cookie.Observed(), cookie.Expected());
// Same thing shrinking it to a single byte, or one byte less than a whole
// page.
ASSERT_TRUE(mapping.ResetAddrLen(orig_addr, 1));
EXPECT_TRUE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), orig_addr);
EXPECT_EQ(mapping.len(), 1u);
EXPECT_EQ(cookie.Observed(), cookie.Expected());
ASSERT_TRUE(mapping.ResetAddrLen(orig_addr, kPageSize - 1));
EXPECT_TRUE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), orig_addr);
EXPECT_EQ(mapping.len(), kPageSize - 1);
EXPECT_EQ(cookie.Observed(), cookie.Expected());
// Shrinking a two-page mapping to a half page frees the second page but
// leaves the first alone.
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, 2 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 2 * kPageSize);
TestCookie two_cookies[2];
for (size_t index = 0; index < base::size(two_cookies); ++index) {
two_cookies[index].SetUp(reinterpret_cast<uint64_t*>(
mapping.addr_as<uintptr_t>() + index * kPageSize));
}
orig_addr = mapping.addr();
ASSERT_TRUE(mapping.ResetAddrLen(orig_addr, kHalfPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), orig_addr);
EXPECT_EQ(mapping.len(), kHalfPageSize);
EXPECT_EQ(two_cookies[0].Observed(), two_cookies[0].Expected());
EXPECT_DEATH_CRASH(two_cookies[1].Check(), "");
// Shrinking a two-page mapping to a page and a half is a no-op.
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, 2 * kPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), 2 * kPageSize);
for (size_t index = 0; index < base::size(two_cookies); ++index) {
two_cookies[index].SetUp(reinterpret_cast<uint64_t*>(
mapping.addr_as<uintptr_t>() + index * kPageSize));
}
orig_addr = mapping.addr();
ASSERT_TRUE(mapping.ResetAddrLen(orig_addr, kPageSize + kHalfPageSize));
EXPECT_TRUE(mapping.is_valid());
EXPECT_EQ(mapping.addr(), orig_addr);
EXPECT_EQ(mapping.len(), kPageSize + kHalfPageSize);
EXPECT_EQ(two_cookies[0].Observed(), two_cookies[0].Expected());
EXPECT_EQ(two_cookies[1].Observed(), two_cookies[1].Expected());
}
TEST(ScopedMmapDeathTest, Mprotect) {
ScopedMmap mapping;
const size_t kPageSize = base::checked_cast<size_t>(getpagesize());
ASSERT_TRUE(ScopedMmapResetMmap(&mapping, kPageSize));
EXPECT_TRUE(mapping.is_valid());
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_NE(mapping.addr(), MAP_FAILED);
EXPECT_EQ(mapping.len(), kPageSize);
char* addr = mapping.addr_as<char*>();
*addr = 1;
ASSERT_TRUE(mapping.Mprotect(PROT_READ));
mac: Tests that crash intentionally shouldn’t go to ReportCrash Crashpad has many tests that crash intentionally. Some of these are gtest death tests, and others arrange for intentional crashes to test Crashpad’s own crash-catching logic. On macOS, all of the gtest death tests and some of the other intentional crashes were being logged by ReportCrash, the system’s crash reporter. Since these reports corresponded to intentional crashes, they were never useful, and served only to clutter ~/Library/Logs/DiagnosticReports. Since Crashpad is adept at handling exceptions on its own, this introduces the “exception swallowing server”, crashpad_exception_swallower, which is a Mach exception server that implements a no-op exception handler routine for all exceptions received. The exception swallowing server is established as the task handler for EXC_CRASH and EXC_CORPSE_NOTIFY exceptions during gtest death tests invoked by {ASSERT,EXPECT}_DEATH_{CHECK,CRASH}, and for all child processes invoked by the Multiprocess test infrastructure. The exception swallowing server is not in effect at other times, so unexpected crashes in test code can still be handled by ReportCrash or another crash reporter. With this change in place, no new reports are generated in the user-level ~/Library/Logs/DiagnosticReports or the system’s /Library/Logs/DiagnosticReports during a run of Crashpad’s full test suite on macOS. Bug: crashpad:33 Change-Id: I13891853a7e25accc30da21fa7ea8bd7d1f3bd2f Reviewed-on: https://chromium-review.googlesource.com/777859 Commit-Queue: Mark Mentovai <mark@chromium.org> Reviewed-by: Robert Sesek <rsesek@chromium.org>
2017-11-20 13:32:26 -05:00
EXPECT_DEATH_CRASH(*addr = 0, "");
ASSERT_TRUE(mapping.Mprotect(PROT_READ | PROT_WRITE));
test: Use (actual, [un]expected) in gtest {ASSERT,EXPECT}_{EQ,NE} gtest used to require (expected, actual) ordering for arguments to EXPECT_EQ and ASSERT_EQ, and in failed test assertions would identify each side as “expected” or “actual.” Tests in Crashpad adhered to this traditional ordering. After a gtest change in February 2016, it is now agnostic with respect to the order of these arguments. This change mechanically updates all uses of these macros to (actual, expected) by reversing them. This provides consistency with our use of the logging CHECK_EQ and DCHECK_EQ macros, and makes for better readability by ordinary native speakers. The rough (but working!) conversion tool is https://chromium-review.googlesource.com/c/466727/1/rewrite_expectassert_eq.py, and “git cl format” cleaned up its output. EXPECT_NE and ASSERT_NE never had a preferred ordering. gtest never made a judgment that one side or the other needed to provide an “unexpected” value. Consequently, some code used (unexpected, actual) while other code used (actual, unexpected). For consistency with the new EXPECT_EQ and ASSERT_EQ usage, as well as consistency with CHECK_NE and DCHECK_NE, this change also updates these use sites to (actual, unexpected) where one side can be called “unexpected” as, for example, std::string::npos can be. Unfortunately, this portion was a manual conversion. References: https://github.com/google/googletest/blob/master/googletest/docs/Primer.md#binary-comparison https://github.com/google/googletest/commit/77d6b173380332b1c1bc540532641f410ec82d65 https://github.com/google/googletest/pull/713 Change-Id: I978fef7c94183b8b1ef63f12f5ab4d6693626be3 Reviewed-on: https://chromium-review.googlesource.com/466727 Reviewed-by: Scott Graham <scottmg@chromium.org>
2017-04-04 00:35:21 -04:00
EXPECT_EQ(*addr, 1);
*addr = 2;
}
} // namespace
} // namespace test
} // namespace crashpad