May 2021 - Linux-kselftest-mirror

x86/fpu/xsave: protection key test failures

by Babu Moger

Hi All, We have been noticing protection key test failures on our systems here. Shggy(Dave Kleikamp) from oracle reported to this problem couple of weeks ago. Here the is the failure. # ./protection_keys_64 has pkeys: 1 startup pkey_reg: 0000000055555550 test 0 PASSED (iteration 1) test 1 PASSED (iteration 1) test 2 PASSED (iteration 1) test 3 PASSED (iteration 1) test 4 PASSED (iteration 1) test 5 PASSED (iteration 1) test 6 PASSED (iteration 1) test 7 PASSED (iteration 1) test 8 PASSED (iteration 1) test 9 PASSED (iteration 1) test 10 PASSED (iteration 1) test 11 PASSED (iteration 1) test 12 PASSED (iteration 1) test 13 PASSED (iteration 1) protection_keys_64: pkey-helpers.h:127: _read_pkey_reg: Assertion `pkey_reg == shadow_pkey_reg' failed. Aborted (core dumped) The test that is failing is "test_ptrace_of_child". Sometimes it fails even earlier also. Sometimes(very rarely) it passes when I insert few printfs. Most probably it fails with test_ptrace_of_child. In the test "test_ptrace_of_child", the parent thread attaches to the child and modifies the data structure associated with protection key. Verifies the access results based on the key permissions. While running the test, the tool finds the key permission changes out of nowhere. The test fails with assert "assert(pkey_reg == shadow_pkey_reg);" Investigation so far. 1. The test fails on AMD(Milan) systems. Test appears to pass on Intel systems. Tested on old Intel system available here. 2. I was trying to see if the hardware(or firmware) is corrupting the pkru register value. At this point, it does not appear that way. I was able to trace all the MSR writes to the application or kernel. At this point, it does not appear to me as an hardware issue. I see that kernel appears to do save/restore properly during the context switch. This value stays default(value 0x55555554) in most cases unless the application changes the default permissions. Only application that changes here is "protection_keys". 3. I played with the test tool little bit. The behavior changes drastically when I make minor changes. For example, in the following code. void setup_handlers(void) { signal(SIGCHLD, &sig_chld); setup_sigsegv_handler(); } Just commenting the first line "signal(SIGCHLD, &sig_chld);" changes the behavior drastically. I see some tests PASS after this change. The first line appear to not do anything except some printing. I still have not figured out completely what is going on with setup_sigsegv_handler();. It seems very odd modifying some structures in the signal handler. I am not sure about some of the offsets it is trying to modify. I feel it may be messing up something there. I am not sure though. Will have to investigate. 4. I took the traces(x86_fpu) while running test. It appears to show some of the feature headers are not copied during the test. But I could not figure out why it was happening. It appears to show not all the feature headers are copied when the new child is created. When I printed the feature bits, they all appear to be intact. Here is the trace. protection_keys-17350 [035] 59275.833511: x86_fpu_regs_activated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834197: x86_fpu_copy_src: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834197: x86_fpu_copy_dst: x86/fpu: 0xffff93d722877800 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17351 [040] 59275.834274: x86_fpu_regs_activated: x86/fpu: 0xffff93d722877800 load: 1 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834283: x86_fpu_regs_deactivated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17351 [040] 59275.834289: x86_fpu_regs_deactivated: x86/fpu: 0xffff93d722877800 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834296: x86_fpu_regs_activated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834298: x86_fpu_regs_activated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834406: x86_fpu_before_save: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 2 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834406: x86_fpu_after_save: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834408: x86_fpu_before_save: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834408: x86_fpu_after_save: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834654: x86_fpu_regs_deactivated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834654: x86_fpu_dropped: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 auditd-1834 [036] 59275.834655: x86_fpu_regs_activated: x86/fpu: 0xffff93d713fef800 load: 1 xfeatures: 202 xcomp_bv: 8000000000000207 protection_keys-17350 [035] 59275.834665: x86_fpu_regs_deactivated: x86/fpu: 0xffff93d7595e2dc0 load: 0 xfeatures: 202 xcomp_bv: 8000000000000207 <...>-17285 [041] 59275.834679: x86_fpu_regs_activated: x86/fpu: 0xffff93d716d0df40 load: 1 xfeatures: 202 xcomp_bv: 8000000000000207 5. I instrumented parent child interactions using a separate standalone application(without the signal handler), it appears to work just fine. I see PRKU register staying intact swiching from parent to child. My suspicion at this point is towards the selftest tool protection_keys.c. I will keep looking. Any feedback would be much appreciated to debug further. Shaggy, Feel free to add if I missed something. Thanks Babu

2 years, 10 months

4
12
0 0

[PATCH v2 00/10] KVM: selftests: exercise userfaultfd minor faults

by Axel Rasmussen

Base ==== These patches are based upon Andrew Morton's v5.13-rc1-mmots-2021-05-13-17-23 tag. This is because this series depends on: - UFFD minor fault support for hugetlbfs (in v5.13-rc1) [1] - UFFD minor fault support for shmem (in Andrew's tree) [2] [1] https://lore.kernel.org/linux-fsdevel/20210301222728.176417-1-axelrasmussen… [2] https://lore.kernel.org/patchwork/cover/1420967/ Changelog ========= v1->v2: - Picked up Reviewed-by's. - Change backing_src_is_shared() to check the flags, instead of the type. This makes it robust to adding new backing source types in the future. - Add another commit which refactors setup_demand_paging() error handling. - Print UFFD ioctl type once in setup_demand_paging, instead of on every page-in operation. - Expand comment on why we use MFD_HUGETLB instead of MAP_HUGETLB. - Reworded comment on addr_gpa2alias. - Moved demand_paging_test.c timing calls outside of the if (), deduplicating them. - Split trivial comment / logging fixups into a separate commit. - Add another commit which prints a clarifying message on test skip. - Split the commit allowing backing src_type to be modified in two. - Split the commit adding the shmem backing type in two. - Rebased onto v5.13-rc1-mmots-2021-05-13-17-23. Overview ======== Minor fault handling is a new userfaultfd feature whose goal is generally to improve performance. In particular, it is intended for use with demand paging. There are more details in the cover letters for this new feature (linked above), but at a high level the idea is that we think of these three phases of live migration of a VM: 1. Precopy, where we copy "some" pages from the source to the target, while the VM is still running on the source machine. 2. Blackout, where execution stops on the source, and begins on the target. 3. Postcopy, where the VM is running on the target, some pages are already up to date, and others are not (because they weren't copied, or were modified after being copied). During postcopy, the first time the guest touches memory, we intercept a minor fault. Userspace checks whether or not the page is already up to date. If needed, we copy the final version of the page from the soure machine. This could be done with RDMA for example, to do it truly in place / with no copying. At this point, all that's left is to setup PTEs for the guest: so we issue UFFDIO_CONTINUE. No copying or page allocation needed. Because of this use case, it's useful to exercise this as part of the demand paging test. It lets us ensure the use case works correctly end-to-end, and also gives us an in-tree way to profile the end-to-end flow for future performance improvements. Axel Rasmussen (10): KVM: selftests: trivial comment/logging fixes KVM: selftests: simplify setup_demand_paging error handling KVM: selftests: print a message when skipping KVM tests KVM: selftests: compute correct demand paging size KVM: selftests: allow different backing source types KVM: selftests: refactor vm_mem_backing_src_type flags KVM: selftests: add shmem backing source type KVM: selftests: create alias mappings when using shared memory KVM: selftests: allow using UFFD minor faults for demand paging KVM: selftests: add shared hugetlbfs backing source type .../selftests/kvm/demand_paging_test.c | 175 +++++++++++------- .../testing/selftests/kvm/include/kvm_util.h | 1 + .../testing/selftests/kvm/include/test_util.h | 12 ++ tools/testing/selftests/kvm/lib/kvm_util.c | 84 ++++++++- .../selftests/kvm/lib/kvm_util_internal.h | 2 + tools/testing/selftests/kvm/lib/test_util.c | 51 +++-- 6 files changed, 238 insertions(+), 87 deletions(-) -- 2.31.1.751.gd2f1c929bd-goog

2 years, 10 months

4
31
0 0

[PATCH] kunit: tool: internal refactor of parser input handling

by Daniel Latypov

Note: this does not change the parser behavior at all (except for making one error message more useful). This is just an internal refactor. The TAP output parser currently operates over a List[str]. This works, but we only ever need to be able to "peek" at the current line and the ability to "pop" it off. Also, using a List means we need to wait for all the output before we can start parsing. While this is not an issue for most tests which are really lightweight, we do have some longer (~5 minutes) tests. This patch introduces an Input wrapper class that * Exposes a peek()/pop() interface instead of manipulating an array * this allows us to more easily add debugging code [1] * Can consume an input from a generator * we can now parse results as tests are running (the parser code currently doesn't print until the end, so no impact yet). * Tracks the current line number to print better error messages * Would allow us to add additional features more easily, e.g. storing N previous lines so we can print out invalid lines in context, etc. [1] The parsing logic is currently quite fragile. E.g. it'll often say the kernel "CRASHED" if there's something slightly wrong with the output format. When debugging a test that had some memory corruption issues, it resulted in very misleading errors from the parser. Now we could easily add this to trace all the lines consumed and why def pop(self) -> str: n = self._next + print(f'popping {n[0]}: {n[1].ljust(40, " ")}| caller={inspect.stack()[1].function}') Example output: popping 77: TAP version 14 | caller=parse_tap_header popping 78: 1..1 | caller=parse_test_plan popping 79: # Subtest: kunit_executor_test | caller=parse_subtest_header popping 80: 1..2 | caller=parse_subtest_plan popping 81: ok 1 - parse_filter_test | caller=parse_ok_not_ok_test_case popping 82: ok 2 - filter_subsuite_test | caller=parse_ok_not_ok_test_case popping 83: ok 1 - kunit_executor_test | caller=parse_ok_not_ok_test_suite If we introduce an invalid line, we can see the parser go down the wrong path: popping 77: TAP version 14 | caller=parse_tap_header popping 78: 1..1 | caller=parse_test_plan popping 79: # Subtest: kunit_executor_test | caller=parse_subtest_header popping 80: 1..2 | caller=parse_subtest_plan popping 81: 1..2 # this is invalid! | caller=parse_ok_not_ok_test_case popping 82: ok 1 - parse_filter_test | caller=parse_ok_not_ok_test_case popping 83: ok 2 - filter_subsuite_test | caller=parse_ok_not_ok_test_case popping 84: ok 1 - kunit_executor_test | caller=parse_ok_not_ok_test_case [ERROR] ran out of lines before end token Signed-off-by: Daniel Latypov <dlatypov(a)google.com> --- tools/testing/kunit/kunit_parser.py | 136 ++++++++++++++++--------- tools/testing/kunit/kunit_tool_test.py | 18 ++-- 2 files changed, 99 insertions(+), 55 deletions(-) diff --git a/tools/testing/kunit/kunit_parser.py b/tools/testing/kunit/kunit_parser.py index e8bcc139702e..65adb386364a 100644 --- a/tools/testing/kunit/kunit_parser.py +++ b/tools/testing/kunit/kunit_parser.py @@ -47,22 +47,63 @@ class TestStatus(Enum): NO_TESTS = auto() FAILURE_TO_PARSE_TESTS = auto() +class Input: + """Provides a more convenient interface over isolate_kunit_output().""" + _lines: Iterator[Tuple[int, str]] + _next: Tuple[int, str] + _done: bool + + def __init__(self, lines: Iterator[Tuple[int, str]]): + self._lines = lines + self._done = False + self._next = (0, '') + self._get_next() + + def _get_next(self) -> None: + try: + self._next = next(self._lines) + except StopIteration: + self._done = True + + def peek(self) -> str: + return self._next[1] + + def pop(self) -> str: + n = self._next + self._get_next() + return n[1] + + def __bool__(self) -> bool: + return not self._done + + # Only used by kunit_tool_test.py. + def __iter__(self) -> Iterator[str]: + while bool(self): + yield self.pop() + + def line_number(self) -> int: + return self._next[0] + kunit_start_re = re.compile(r'TAP version [0-9]+$') kunit_end_re = re.compile('(List of all partitions:|' 'Kernel panic - not syncing: VFS:)') -def isolate_kunit_output(kernel_output) -> Iterator[str]: - started = False - for line in kernel_output: - line = line.rstrip() # line always has a trailing \n - if kunit_start_re.search(line): - prefix_len = len(line.split('TAP version')[0]) - started = True - yield line[prefix_len:] if prefix_len > 0 else line - elif kunit_end_re.search(line): - break - elif started: - yield line[prefix_len:] if prefix_len > 0 else line +def get_input(kernel_output: Iterable[str]) -> Input: + def isolate_kunit_output(kernel_output: Iterable[str]) -> Iterator[Tuple[int, str]]: + line_num = 0 + started = False + for line in kernel_output: + line_num += 1 + line = line.rstrip() # line always has a trailing \n + if kunit_start_re.search(line): + prefix_len = len(line.split('TAP version')[0]) + started = True + yield line_num, line[prefix_len:] + elif kunit_end_re.search(line): + break + elif started: + yield line_num, line[prefix_len:] + return Input(lines=isolate_kunit_output(kernel_output)) def raw_output(kernel_output) -> None: for line in kernel_output: @@ -97,14 +138,14 @@ def print_log(log) -> None: TAP_ENTRIES = re.compile(r'^(TAP|[\s]*ok|[\s]*not ok|[\s]*[0-9]+\.\.[0-9]+|[\s]*#).*$') -def consume_non_diagnostic(lines: List[str]) -> None: - while lines and not TAP_ENTRIES.match(lines[0]): - lines.pop(0) +def consume_non_diagnostic(lines: Input) -> None: + while lines and not TAP_ENTRIES.match(lines.peek()): + lines.pop() -def save_non_diagnostic(lines: List[str], test_case: TestCase) -> None: - while lines and not TAP_ENTRIES.match(lines[0]): - test_case.log.append(lines[0]) - lines.pop(0) +def save_non_diagnostic(lines: Input, test_case: TestCase) -> None: + while lines and not TAP_ENTRIES.match(lines.peek()): + test_case.log.append(lines.peek()) + lines.pop() OkNotOkResult = namedtuple('OkNotOkResult', ['is_ok','description', 'text']) @@ -112,18 +153,18 @@ OK_NOT_OK_SUBTEST = re.compile(r'^[\s]+(ok|not ok) [0-9]+ - (.*)$') OK_NOT_OK_MODULE = re.compile(r'^(ok|not ok) ([0-9]+) - (.*)$') -def parse_ok_not_ok_test_case(lines: List[str], test_case: TestCase) -> bool: +def parse_ok_not_ok_test_case(lines: Input, test_case: TestCase) -> bool: save_non_diagnostic(lines, test_case) if not lines: test_case.status = TestStatus.TEST_CRASHED return True - line = lines[0] + line = lines.peek() match = OK_NOT_OK_SUBTEST.match(line) while not match and lines: - line = lines.pop(0) + line = lines.pop() match = OK_NOT_OK_SUBTEST.match(line) if match: - test_case.log.append(lines.pop(0)) + test_case.log.append(lines.pop()) test_case.name = match.group(2) if test_case.status == TestStatus.TEST_CRASHED: return True @@ -138,14 +179,14 @@ def parse_ok_not_ok_test_case(lines: List[str], test_case: TestCase) -> bool: SUBTEST_DIAGNOSTIC = re.compile(r'^[\s]+# (.*)$') DIAGNOSTIC_CRASH_MESSAGE = re.compile(r'^[\s]+# .*?: kunit test case crashed!$') -def parse_diagnostic(lines: List[str], test_case: TestCase) -> bool: +def parse_diagnostic(lines: Input, test_case: TestCase) -> bool: save_non_diagnostic(lines, test_case) if not lines: return False - line = lines[0] + line = lines.peek() match = SUBTEST_DIAGNOSTIC.match(line) if match: - test_case.log.append(lines.pop(0)) + test_case.log.append(lines.pop()) crash_match = DIAGNOSTIC_CRASH_MESSAGE.match(line) if crash_match: test_case.status = TestStatus.TEST_CRASHED @@ -153,7 +194,7 @@ def parse_diagnostic(lines: List[str], test_case: TestCase) -> bool: else: return False -def parse_test_case(lines: List[str]) -> Optional[TestCase]: +def parse_test_case(lines: Input) -> Optional[TestCase]: test_case = TestCase() save_non_diagnostic(lines, test_case) while parse_diagnostic(lines, test_case): @@ -165,24 +206,24 @@ def parse_test_case(lines: List[str]) -> Optional[TestCase]: SUBTEST_HEADER = re.compile(r'^[\s]+# Subtest: (.*)$') -def parse_subtest_header(lines: List[str]) -> Optional[str]: +def parse_subtest_header(lines: Input) -> Optional[str]: consume_non_diagnostic(lines) if not lines: return None - match = SUBTEST_HEADER.match(lines[0]) + match = SUBTEST_HEADER.match(lines.peek()) if match: - lines.pop(0) + lines.pop() return match.group(1) else: return None SUBTEST_PLAN = re.compile(r'[\s]+[0-9]+\.\.([0-9]+)') -def parse_subtest_plan(lines: List[str]) -> Optional[int]: +def parse_subtest_plan(lines: Input) -> Optional[int]: consume_non_diagnostic(lines) - match = SUBTEST_PLAN.match(lines[0]) + match = SUBTEST_PLAN.match(lines.peek()) if match: - lines.pop(0) + lines.pop() return int(match.group(1)) else: return None @@ -199,17 +240,17 @@ def max_status(left: TestStatus, right: TestStatus) -> TestStatus: else: return TestStatus.SUCCESS -def parse_ok_not_ok_test_suite(lines: List[str], +def parse_ok_not_ok_test_suite(lines: Input, test_suite: TestSuite, expected_suite_index: int) -> bool: consume_non_diagnostic(lines) if not lines: test_suite.status = TestStatus.TEST_CRASHED return False - line = lines[0] + line = lines.peek() match = OK_NOT_OK_MODULE.match(line) if match: - lines.pop(0) + lines.pop() if match.group(1) == 'ok': test_suite.status = TestStatus.SUCCESS else: @@ -231,7 +272,7 @@ def bubble_up_test_case_errors(test_suite: TestSuite) -> TestStatus: max_test_case_status = bubble_up_errors(x.status for x in test_suite.cases) return max_status(max_test_case_status, test_suite.status) -def parse_test_suite(lines: List[str], expected_suite_index: int) -> Optional[TestSuite]: +def parse_test_suite(lines: Input, expected_suite_index: int) -> Optional[TestSuite]: if not lines: return None consume_non_diagnostic(lines) @@ -257,26 +298,26 @@ def parse_test_suite(lines: List[str], expected_suite_index: int) -> Optional[Te print_with_timestamp(red('[ERROR] ') + 'ran out of lines before end token') return test_suite else: - print('failed to parse end of suite' + lines[0]) + print(f'failed to parse end of suite "{name}", at line {lines.line_number()}: {lines.peek()}') return None TAP_HEADER = re.compile(r'^TAP version 14$') -def parse_tap_header(lines: List[str]) -> bool: +def parse_tap_header(lines: Input) -> bool: consume_non_diagnostic(lines) - if TAP_HEADER.match(lines[0]): - lines.pop(0) + if TAP_HEADER.match(lines.peek()): + lines.pop() return True else: return False TEST_PLAN = re.compile(r'[0-9]+\.\.([0-9]+)') -def parse_test_plan(lines: List[str]) -> Optional[int]: +def parse_test_plan(lines: Input) -> Optional[int]: consume_non_diagnostic(lines) - match = TEST_PLAN.match(lines[0]) + match = TEST_PLAN.match(lines.peek()) if match: - lines.pop(0) + lines.pop() return int(match.group(1)) else: return None @@ -284,7 +325,7 @@ def parse_test_plan(lines: List[str]) -> Optional[int]: def bubble_up_suite_errors(test_suites: Iterable[TestSuite]) -> TestStatus: return bubble_up_errors(x.status for x in test_suites) -def parse_test_result(lines: List[str]) -> TestResult: +def parse_test_result(lines: Input) -> TestResult: consume_non_diagnostic(lines) if not lines or not parse_tap_header(lines): return TestResult(TestStatus.NO_TESTS, [], lines) @@ -338,11 +379,12 @@ def print_and_count_results(test_result: TestResult) -> Tuple[int, int, int]: print_with_timestamp('') return total_tests, failed_tests, crashed_tests -def parse_run_tests(kernel_output) -> TestResult: +def parse_run_tests(kernel_output: Iterable[str]) -> TestResult: total_tests = 0 failed_tests = 0 crashed_tests = 0 - test_result = parse_test_result(list(isolate_kunit_output(kernel_output))) + lines = get_input(kernel_output) + test_result = parse_test_result(lines) if test_result.status == TestStatus.NO_TESTS: print(red('[ERROR] ') + yellow('no tests run!')) elif test_result.status == TestStatus.FAILURE_TO_PARSE_TESTS: diff --git a/tools/testing/kunit/kunit_tool_test.py b/tools/testing/kunit/kunit_tool_test.py index 2e809dd956a7..e82678a25bef 100755 --- a/tools/testing/kunit/kunit_tool_test.py +++ b/tools/testing/kunit/kunit_tool_test.py @@ -11,6 +11,7 @@ from unittest import mock import tempfile, shutil # Handling test_tmpdir +import itertools import json import signal import os @@ -92,17 +93,18 @@ class KconfigTest(unittest.TestCase): class KUnitParserTest(unittest.TestCase): - def assertContains(self, needle, haystack): - for line in haystack: + def assertContains(self, needle: str, haystack: kunit_parser.Input): + # Clone the iterator so we can print the contents on failure. + copy, backup = itertools.tee(haystack) + for line in copy: if needle in line: return - raise AssertionError('"' + - str(needle) + '" not found in "' + str(haystack) + '"!') + raise AssertionError(f'"{needle}" not found in {list(backup)}!') def test_output_isolated_correctly(self): log_path = test_data_path('test_output_isolated_correctly.log') with open(log_path) as file: - result = kunit_parser.isolate_kunit_output(file.readlines()) + result = kunit_parser.get_input(file.readlines()) self.assertContains('TAP version 14', result) self.assertContains(' # Subtest: example', result) self.assertContains(' 1..2', result) @@ -113,7 +115,7 @@ class KUnitParserTest(unittest.TestCase): def test_output_with_prefix_isolated_correctly(self): log_path = test_data_path('test_pound_sign.log') with open(log_path) as file: - result = kunit_parser.isolate_kunit_output(file.readlines()) + result = kunit_parser.get_input(file.readlines()) self.assertContains('TAP version 14', result) self.assertContains(' # Subtest: kunit-resource-test', result) self.assertContains(' 1..5', result) @@ -159,7 +161,7 @@ class KUnitParserTest(unittest.TestCase): empty_log = test_data_path('test_is_test_passed-no_tests_run.log') with open(empty_log) as file: result = kunit_parser.parse_run_tests( - kunit_parser.isolate_kunit_output(file.readlines())) + kunit_parser.get_input(file.readlines())) self.assertEqual(0, len(result.suites)) self.assertEqual( kunit_parser.TestStatus.NO_TESTS, @@ -170,7 +172,7 @@ class KUnitParserTest(unittest.TestCase): print_mock = mock.patch('builtins.print').start() with open(crash_log) as file: result = kunit_parser.parse_run_tests( - kunit_parser.isolate_kunit_output(file.readlines())) + kunit_parser.get_input(file.readlines())) print_mock.assert_any_call(StrContains('no tests run!')) print_mock.stop() file.close() base-commit: c3d0e3fd41b7f0f5d5d5b6022ab7e813f04ea727 -- 2.31.1.818.g46aad6cb9e-goog

2 years, 10 months

2
2
0 0

[PATCH v6 1/3] selftests/sgx: Rename 'eenter' and 'sgx_call_vdso'

by Jarkko Sakkinen

Rename symbols for better clarity: * 'eenter' might be confused for directly calling ENCLU[EENTER]. It does not. It calls into the VDSO, which actually has the EENTER instruction. * 'sgx_call_vdso' is *only* used for entering the enclave. It's not some generic SGX call into the VDSO. Signed-off-by: Jarkko Sakkinen <jarkko(a)kernel.org> --- v3: Refine the commit message according to Dave's suggestions. v2: Refined the renames just a bit. tools/testing/selftests/sgx/call.S | 6 +++--- tools/testing/selftests/sgx/main.c | 25 +++++++++++++------------ tools/testing/selftests/sgx/main.h | 4 ++-- 3 files changed, 18 insertions(+), 17 deletions(-) diff --git a/tools/testing/selftests/sgx/call.S b/tools/testing/selftests/sgx/call.S index 4ecadc7490f4..b09a25890f3b 100644 --- a/tools/testing/selftests/sgx/call.S +++ b/tools/testing/selftests/sgx/call.S @@ -5,8 +5,8 @@ .text - .global sgx_call_vdso -sgx_call_vdso: + .global sgx_enter_enclave +sgx_enter_enclave: .cfi_startproc push %r15 .cfi_adjust_cfa_offset 8 @@ -27,7 +27,7 @@ sgx_call_vdso: .cfi_adjust_cfa_offset 8 push 0x38(%rsp) .cfi_adjust_cfa_offset 8 - call *eenter(%rip) + call *vdso_sgx_enter_enclave(%rip) add $0x10, %rsp .cfi_adjust_cfa_offset -0x10 pop %rbx diff --git a/tools/testing/selftests/sgx/main.c b/tools/testing/selftests/sgx/main.c index d304a4044eb9..43da68388e25 100644 --- a/tools/testing/selftests/sgx/main.c +++ b/tools/testing/selftests/sgx/main.c @@ -21,7 +21,7 @@ #include "../kselftest.h" static const uint64_t MAGIC = 0x1122334455667788ULL; -vdso_sgx_enter_enclave_t eenter; +vdso_sgx_enter_enclave_t vdso_sgx_enter_enclave; struct vdso_symtab { Elf64_Sym *elf_symtab; @@ -149,7 +149,7 @@ int main(int argc, char *argv[]) { struct sgx_enclave_run run; struct vdso_symtab symtab; - Elf64_Sym *eenter_sym; + Elf64_Sym *sgx_enter_enclave_sym; uint64_t result = 0; struct encl encl; unsigned int i; @@ -194,29 +194,30 @@ int main(int argc, char *argv[]) if (!vdso_get_symtab(addr, &symtab)) goto err; - eenter_sym = vdso_symtab_get(&symtab, "__vdso_sgx_enter_enclave"); - if (!eenter_sym) + sgx_enter_enclave_sym = vdso_symtab_get(&symtab, "__vdso_sgx_enter_enclave"); + if (!sgx_enter_enclave_sym) goto err; - eenter = addr + eenter_sym->st_value; + vdso_sgx_enter_enclave = addr + sgx_enter_enclave_sym->st_value; - ret = sgx_call_vdso((void *)&MAGIC, &result, 0, EENTER, NULL, NULL, &run); - if (!report_results(&run, ret, result, "sgx_call_vdso")) + ret = sgx_enter_enclave((void *)&MAGIC, &result, 0, EENTER, + NULL, NULL, &run); + if (!report_results(&run, ret, result, "sgx_enter_enclave_unclobbered")) goto err; /* Invoke the vDSO directly. */ result = 0; - ret = eenter((unsigned long)&MAGIC, (unsigned long)&result, 0, EENTER, - 0, 0, &run); - if (!report_results(&run, ret, result, "eenter")) + ret = vdso_sgx_enter_enclave((unsigned long)&MAGIC, (unsigned long)&result, + 0, EENTER, 0, 0, &run); + if (!report_results(&run, ret, result, "sgx_enter_enclave")) goto err; /* And with an exit handler. */ run.user_handler = (__u64)user_handler; run.user_data = 0xdeadbeef; - ret = eenter((unsigned long)&MAGIC, (unsigned long)&result, 0, EENTER, - 0, 0, &run); + ret = vdso_sgx_enter_enclave((unsigned long)&MAGIC, (unsigned long)&result, + 0, EENTER, 0, 0, &run); if (!report_results(&run, ret, result, "user_handler")) goto err; diff --git a/tools/testing/selftests/sgx/main.h b/tools/testing/selftests/sgx/main.h index 67211a708f04..68672fd86cf9 100644 --- a/tools/testing/selftests/sgx/main.h +++ b/tools/testing/selftests/sgx/main.h @@ -35,7 +35,7 @@ bool encl_load(const char *path, struct encl *encl); bool encl_measure(struct encl *encl); bool encl_build(struct encl *encl); -int sgx_call_vdso(void *rdi, void *rsi, long rdx, u32 function, void *r8, void *r9, - struct sgx_enclave_run *run); +int sgx_enter_enclave(void *rdi, void *rsi, long rdx, u32 function, void *r8, void *r9, + struct sgx_enclave_run *run); #endif /* MAIN_H */ -- 2.31.1

2 years, 10 months

1
2
0 0

[PATCH] kselftest/arm64: Add missing newline to SVE test skipping output

by Mark Brown

The newline is expected to come from the caller but got missed for this test. Signed-off-by: Mark Brown <broonie(a)kernel.org> --- tools/testing/selftests/arm64/fp/sve-probe-vls.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/arm64/fp/sve-probe-vls.c b/tools/testing/selftests/arm64/fp/sve-probe-vls.c index b29cbc642c57..76e138525d55 100644 --- a/tools/testing/selftests/arm64/fp/sve-probe-vls.c +++ b/tools/testing/selftests/arm64/fp/sve-probe-vls.c @@ -25,7 +25,7 @@ int main(int argc, char **argv) ksft_set_plan(2); if (!(getauxval(AT_HWCAP) & HWCAP_SVE)) - ksft_exit_skip("SVE not available"); + ksft_exit_skip("SVE not available\n"); /* * Enumerate up to SVE_VQ_MAX vector lengths -- 2.20.1

2 years, 10 months

2
1
0 0

[PATCH] selftests: Use kselftest skip code for skipped tests

by Po-Hsu Lin

There are several test cases still using exit 0 when they need to be skipped. Use kselftest framework skip code instead so it can help us to distinguish the proper return status. Criterion to filter out what should be fixed in selftests directory: grep -r "exit 0" -B1 | grep -i skip This change might cause some false-positives if people are running these test scripts directly and only checking their return codes, which will change from 0 to 4. However I think the impact should be small as most of our scripts here are already using this skip code. And there will be no such issue if running them with the kselftest framework. Signed-off-by: Po-Hsu Lin <po-hsu.lin(a)canonical.com> --- tools/testing/selftests/bpf/test_bpftool_build.sh | 5 ++++- tools/testing/selftests/bpf/test_xdp_meta.sh | 5 ++++- tools/testing/selftests/bpf/test_xdp_vlan.sh | 7 +++++-- tools/testing/selftests/net/fcnal-test.sh | 5 ++++- tools/testing/selftests/net/fib_rule_tests.sh | 7 +++++-- tools/testing/selftests/net/forwarding/lib.sh | 5 ++++- tools/testing/selftests/net/forwarding/router_mpath_nh.sh | 5 ++++- tools/testing/selftests/net/forwarding/router_mpath_nh_res.sh | 5 ++++- tools/testing/selftests/net/run_afpackettests | 5 ++++- tools/testing/selftests/net/srv6_end_dt4_l3vpn_test.sh | 9 ++++++--- tools/testing/selftests/net/srv6_end_dt6_l3vpn_test.sh | 9 ++++++--- tools/testing/selftests/net/unicast_extensions.sh | 5 ++++- tools/testing/selftests/net/vrf_strict_mode_test.sh | 9 ++++++--- tools/testing/selftests/ptp/phc.sh | 7 +++++-- tools/testing/selftests/vm/charge_reserved_hugetlb.sh | 5 ++++- tools/testing/selftests/vm/hugetlb_reparenting_test.sh | 5 ++++- 16 files changed, 73 insertions(+), 25 deletions(-) diff --git a/tools/testing/selftests/bpf/test_bpftool_build.sh b/tools/testing/selftests/bpf/test_bpftool_build.sh index ac349a5..b6fab1e 100755 --- a/tools/testing/selftests/bpf/test_bpftool_build.sh +++ b/tools/testing/selftests/bpf/test_bpftool_build.sh @@ -1,6 +1,9 @@ #!/bin/bash # SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + case $1 in -h|--help) echo -e "$0 [-j <n>]" @@ -22,7 +25,7 @@ KDIR_ROOT_DIR=$(realpath $PWD/$SCRIPT_REL_DIR/../../../../) cd $KDIR_ROOT_DIR if [ ! -e tools/bpf/bpftool/Makefile ]; then echo -e "skip: bpftool files not found!\n" - exit 0 + exit $ksft_skip fi ERROR=0 diff --git a/tools/testing/selftests/bpf/test_xdp_meta.sh b/tools/testing/selftests/bpf/test_xdp_meta.sh index 637fcf4..fd3f218 100755 --- a/tools/testing/selftests/bpf/test_xdp_meta.sh +++ b/tools/testing/selftests/bpf/test_xdp_meta.sh @@ -1,5 +1,8 @@ #!/bin/sh +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + cleanup() { if [ "$?" = "0" ]; then @@ -17,7 +20,7 @@ cleanup() ip link set dev lo xdp off 2>/dev/null > /dev/null if [ $? -ne 0 ];then echo "selftests: [SKIP] Could not run test without the ip xdp support" - exit 0 + exit $ksft_skip fi set -e diff --git a/tools/testing/selftests/bpf/test_xdp_vlan.sh b/tools/testing/selftests/bpf/test_xdp_vlan.sh index bb8b0da..1aa7404 100755 --- a/tools/testing/selftests/bpf/test_xdp_vlan.sh +++ b/tools/testing/selftests/bpf/test_xdp_vlan.sh @@ -2,6 +2,9 @@ # SPDX-License-Identifier: GPL-2.0 # Author: Jesper Dangaard Brouer <hawk(a)kernel.org> +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + # Allow wrapper scripts to name test if [ -z "$TESTNAME" ]; then TESTNAME=xdp_vlan @@ -94,7 +97,7 @@ while true; do -h | --help ) usage; echo "selftests: $TESTNAME [SKIP] usage help info requested" - exit 0 + exit $ksft_skip ;; * ) shift @@ -117,7 +120,7 @@ fi ip link set dev lo xdpgeneric off 2>/dev/null > /dev/null if [ $? -ne 0 ]; then echo "selftests: $TESTNAME [SKIP] need ip xdp support" - exit 0 + exit $ksft_skip fi # Interactive mode likely require us to cleanup netns diff --git a/tools/testing/selftests/net/fcnal-test.sh b/tools/testing/selftests/net/fcnal-test.sh index a8ad928..9074e25 100755 --- a/tools/testing/selftests/net/fcnal-test.sh +++ b/tools/testing/selftests/net/fcnal-test.sh @@ -37,6 +37,9 @@ # # server / client nomenclature relative to ns-A +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + VERBOSE=0 NSA_DEV=eth1 @@ -3946,7 +3949,7 @@ fi which nettest >/dev/null if [ $? -ne 0 ]; then echo "'nettest' command not found; skipping tests" - exit 0 + exit $ksft_skip fi declare -i nfail=0 diff --git a/tools/testing/selftests/net/fib_rule_tests.sh b/tools/testing/selftests/net/fib_rule_tests.sh index a93e6b6..43ea840 100755 --- a/tools/testing/selftests/net/fib_rule_tests.sh +++ b/tools/testing/selftests/net/fib_rule_tests.sh @@ -3,6 +3,9 @@ # This test is for checking IPv4 and IPv6 FIB rules API +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + ret=0 PAUSE_ON_FAIL=${PAUSE_ON_FAIL:=no} @@ -238,12 +241,12 @@ run_fibrule_tests() if [ "$(id -u)" -ne 0 ];then echo "SKIP: Need root privileges" - exit 0 + exit $ksft_skip fi if [ ! -x "$(command -v ip)" ]; then echo "SKIP: Could not run test without ip tool" - exit 0 + exit $ksft_skip fi # start clean diff --git a/tools/testing/selftests/net/forwarding/lib.sh b/tools/testing/selftests/net/forwarding/lib.sh index 42e28c9..eed9f08 100644 --- a/tools/testing/selftests/net/forwarding/lib.sh +++ b/tools/testing/selftests/net/forwarding/lib.sh @@ -4,6 +4,9 @@ ############################################################################## # Defines +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + # Can be overridden by the configuration file. PING=${PING:=ping} PING6=${PING6:=ping6} @@ -121,7 +124,7 @@ check_ethtool_lanes_support() if [[ "$(id -u)" -ne 0 ]]; then echo "SKIP: need root privileges" - exit 0 + exit $ksft_skip fi if [[ "$CHECK_TC" = "yes" ]]; then diff --git a/tools/testing/selftests/net/forwarding/router_mpath_nh.sh b/tools/testing/selftests/net/forwarding/router_mpath_nh.sh index 76efb1f..bb7dc6d 100755 --- a/tools/testing/selftests/net/forwarding/router_mpath_nh.sh +++ b/tools/testing/selftests/net/forwarding/router_mpath_nh.sh @@ -1,6 +1,9 @@ #!/bin/bash # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + ALL_TESTS=" ping_ipv4 ping_ipv6 @@ -411,7 +414,7 @@ ping_ipv6() ip nexthop ls >/dev/null 2>&1 if [ $? -ne 0 ]; then echo "Nexthop objects not supported; skipping tests" - exit 0 + exit $ksft_skip fi trap cleanup EXIT diff --git a/tools/testing/selftests/net/forwarding/router_mpath_nh_res.sh b/tools/testing/selftests/net/forwarding/router_mpath_nh_res.sh index 4898dd4..e7bb976 100755 --- a/tools/testing/selftests/net/forwarding/router_mpath_nh_res.sh +++ b/tools/testing/selftests/net/forwarding/router_mpath_nh_res.sh @@ -1,6 +1,9 @@ #!/bin/bash # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + ALL_TESTS=" ping_ipv4 ping_ipv6 @@ -386,7 +389,7 @@ ping_ipv6() ip nexthop ls >/dev/null 2>&1 if [ $? -ne 0 ]; then echo "Nexthop objects not supported; skipping tests" - exit 0 + exit $ksft_skip fi trap cleanup EXIT diff --git a/tools/testing/selftests/net/run_afpackettests b/tools/testing/selftests/net/run_afpackettests index 8b42e8b..a59cb6a 100755 --- a/tools/testing/selftests/net/run_afpackettests +++ b/tools/testing/selftests/net/run_afpackettests @@ -1,9 +1,12 @@ #!/bin/sh # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + if [ $(id -u) != 0 ]; then echo $msg must be run as root >&2 - exit 0 + exit $ksft_skip fi ret=0 diff --git a/tools/testing/selftests/net/srv6_end_dt4_l3vpn_test.sh b/tools/testing/selftests/net/srv6_end_dt4_l3vpn_test.sh index ad7a9fc..1003119 100755 --- a/tools/testing/selftests/net/srv6_end_dt4_l3vpn_test.sh +++ b/tools/testing/selftests/net/srv6_end_dt4_l3vpn_test.sh @@ -163,6 +163,9 @@ # +---------------------------------------------------+ # +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + readonly LOCALSID_TABLE_ID=90 readonly IPv6_RT_NETWORK=fd00 readonly IPv4_HS_NETWORK=10.0.0 @@ -464,18 +467,18 @@ host_vpn_isolation_tests() if [ "$(id -u)" -ne 0 ];then echo "SKIP: Need root privileges" - exit 0 + exit $ksft_skip fi if [ ! -x "$(command -v ip)" ]; then echo "SKIP: Could not run test without ip tool" - exit 0 + exit $ksft_skip fi modprobe vrf &>/dev/null if [ ! -e /proc/sys/net/vrf/strict_mode ]; then echo "SKIP: vrf sysctl does not exist" - exit 0 + exit $ksft_skip fi cleanup &>/dev/null diff --git a/tools/testing/selftests/net/srv6_end_dt6_l3vpn_test.sh b/tools/testing/selftests/net/srv6_end_dt6_l3vpn_test.sh index 68708f5..b9b06ef 100755 --- a/tools/testing/selftests/net/srv6_end_dt6_l3vpn_test.sh +++ b/tools/testing/selftests/net/srv6_end_dt6_l3vpn_test.sh @@ -164,6 +164,9 @@ # +---------------------------------------------------+ # +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + readonly LOCALSID_TABLE_ID=90 readonly IPv6_RT_NETWORK=fd00 readonly IPv6_HS_NETWORK=cafe @@ -472,18 +475,18 @@ host_vpn_isolation_tests() if [ "$(id -u)" -ne 0 ];then echo "SKIP: Need root privileges" - exit 0 + exit $ksft_skip fi if [ ! -x "$(command -v ip)" ]; then echo "SKIP: Could not run test without ip tool" - exit 0 + exit $ksft_skip fi modprobe vrf &>/dev/null if [ ! -e /proc/sys/net/vrf/strict_mode ]; then echo "SKIP: vrf sysctl does not exist" - exit 0 + exit $ksft_skip fi cleanup &>/dev/null diff --git a/tools/testing/selftests/net/unicast_extensions.sh b/tools/testing/selftests/net/unicast_extensions.sh index dbf0421..728e4d5 100755 --- a/tools/testing/selftests/net/unicast_extensions.sh +++ b/tools/testing/selftests/net/unicast_extensions.sh @@ -28,12 +28,15 @@ # These tests provide an easy way to flip the expected result of any # of these behaviors for testing kernel patches that change them. +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + # nettest can be run from PATH or from same directory as this selftest if ! which nettest >/dev/null; then PATH=$PWD:$PATH if ! which nettest >/dev/null; then echo "'nettest' command not found; skipping tests" - exit 0 + exit $ksft_skip fi fi diff --git a/tools/testing/selftests/net/vrf_strict_mode_test.sh b/tools/testing/selftests/net/vrf_strict_mode_test.sh index 18b982d..865d53c 100755 --- a/tools/testing/selftests/net/vrf_strict_mode_test.sh +++ b/tools/testing/selftests/net/vrf_strict_mode_test.sh @@ -3,6 +3,9 @@ # This test is designed for testing the new VRF strict_mode functionality. +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + ret=0 # identifies the "init" network namespace which is often called root network @@ -371,18 +374,18 @@ vrf_strict_mode_check_support() if [ "$(id -u)" -ne 0 ];then echo "SKIP: Need root privileges" - exit 0 + exit $ksft_skip fi if [ ! -x "$(command -v ip)" ]; then echo "SKIP: Could not run test without ip tool" - exit 0 + exit $ksft_skip fi modprobe vrf &>/dev/null if [ ! -e /proc/sys/net/vrf/strict_mode ]; then echo "SKIP: vrf sysctl does not exist" - exit 0 + exit $ksft_skip fi cleanup &> /dev/null diff --git a/tools/testing/selftests/ptp/phc.sh b/tools/testing/selftests/ptp/phc.sh index ac6e5a6..ca3c844c 100755 --- a/tools/testing/selftests/ptp/phc.sh +++ b/tools/testing/selftests/ptp/phc.sh @@ -1,6 +1,9 @@ #!/bin/bash # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + ALL_TESTS=" settime adjtime @@ -13,12 +16,12 @@ DEV=$1 if [[ "$(id -u)" -ne 0 ]]; then echo "SKIP: need root privileges" - exit 0 + exit $ksft_skip fi if [[ "$DEV" == "" ]]; then echo "SKIP: PTP device not provided" - exit 0 + exit $ksft_skip fi require_command() diff --git a/tools/testing/selftests/vm/charge_reserved_hugetlb.sh b/tools/testing/selftests/vm/charge_reserved_hugetlb.sh index 18d3368..fe8fcfb 100644 --- a/tools/testing/selftests/vm/charge_reserved_hugetlb.sh +++ b/tools/testing/selftests/vm/charge_reserved_hugetlb.sh @@ -1,11 +1,14 @@ #!/bin/sh # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + set -e if [[ $(id -u) -ne 0 ]]; then echo "This test must be run as root. Skipping..." - exit 0 + exit $ksft_skip fi fault_limit_file=limit_in_bytes diff --git a/tools/testing/selftests/vm/hugetlb_reparenting_test.sh b/tools/testing/selftests/vm/hugetlb_reparenting_test.sh index d11d1fe..4a9a3af 100644 --- a/tools/testing/selftests/vm/hugetlb_reparenting_test.sh +++ b/tools/testing/selftests/vm/hugetlb_reparenting_test.sh @@ -1,11 +1,14 @@ #!/bin/bash # SPDX-License-Identifier: GPL-2.0 +# Kselftest framework requirement - SKIP code is 4. +ksft_skip=4 + set -e if [[ $(id -u) -ne 0 ]]; then echo "This test must be run as root. Skipping..." - exit 0 + exit $ksft_skip fi usage_file=usage_in_bytes -- 2.7.4

2 years, 10 months

3
3
0 0

[PATCH] selftests: kvm: make allocation of extra memory take effect

by Zhenzhong Duan

The extra memory pages is missed to be allocated during VM creating. perf_test_util and kvm_page_table_test use it to alloc extra memory currently. Fix it by adding extra_mem_pages to the total memory calculation before allocate. Signed-off-by: Zhenzhong Duan <zhenzhong.duan(a)intel.com> --- tools/testing/selftests/kvm/lib/kvm_util.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c index fc83f6c5902d..159f4d62241d 100644 --- a/tools/testing/selftests/kvm/lib/kvm_util.c +++ b/tools/testing/selftests/kvm/lib/kvm_util.c @@ -295,7 +295,7 @@ struct kvm_vm *vm_create_with_vcpus(enum vm_guest_mode mode, uint32_t nr_vcpus, */ uint64_t vcpu_pages = (DEFAULT_STACK_PGS + num_percpu_pages) * nr_vcpus; uint64_t extra_pg_pages = (extra_mem_pages + vcpu_pages) / PTES_PER_MIN_PAGE * 2; - uint64_t pages = DEFAULT_GUEST_PHY_PAGES + vcpu_pages + extra_pg_pages; + uint64_t pages = DEFAULT_GUEST_PHY_PAGES + extra_mem_pages + vcpu_pages + extra_pg_pages; struct kvm_vm *vm; int i; -- 2.25.1

2 years, 11 months

4
4
0 0

[PATCH 1/2] KVM: Deliver VM fault signals to userspace

by James Houghton

This patch has been written to support page-ins using userfaultfd's SIGBUS feature. When a userfaultfd is created with UFFD_FEATURE_SIGBUS, `handle_userfault` will return VM_FAULT_SIGBUS instead of putting the calling thread to sleep. Normal (non-guest) threads that access memory that has been registered with a UFFD_FEATURE_SIGBUS userfaultfd receive a SIGBUS. When a vCPU gets an EPT page fault in a userfaultfd-registered region, KVM calls into `handle_userfault` to resolve the page fault. With UFFD_FEATURE_SIGBUS, VM_FAULT_SIGBUS is returned, but a SIGBUS is never delivered to the userspace thread. This patch propagates the VM_FAULT_SIGBUS error up to KVM, where we then send the signal. Upon receiving a VM_FAULT_SIGBUS, the KVM_RUN ioctl will exit to userspace. This functionality already exists. This allows a hypervisor to do page-ins with UFFD_FEATURE_SIGBUS: 1. Setup a SIGBUS handler to save the address of the SIGBUS (to a thread-local variable). 2. Enter the guest. 3. Immediately after KVM_RUN returns, check if the address has been set. 4. If an address has been set, we exited due to a page fault that we can now handle. 5. Userspace can do anything it wants to make the memory available, using MODE_NOWAKE for the UFFDIO memory installation ioctls. 6. Re-enter the guest. If the memory still isn't ready, this process will repeat. This style of demand paging is significantly faster than the standard poll/read/wake mechanism userfaultfd uses and completely bypasses the userfaultfd waitq. For a single vCPU, page-in throughput increases by about 3-4x. Signed-off-by: James Houghton <jthoughton(a)google.com> Suggested-by: Jue Wang <juew(a)google.com> --- include/linux/hugetlb.h | 2 +- include/linux/mm.h | 3 ++- mm/gup.c | 57 +++++++++++++++++++++++++++-------------- mm/hugetlb.c | 5 +++- virt/kvm/kvm_main.c | 30 +++++++++++++++++++++- 5 files changed, 74 insertions(+), 23 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index b92f25ccef58..a777fb254df0 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -119,7 +119,7 @@ int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *, struct vm_ar long follow_hugetlb_page(struct mm_struct *, struct vm_area_struct *, struct page **, struct vm_area_struct **, unsigned long *, unsigned long *, long, unsigned int, - int *); + int *, int *); void unmap_hugepage_range(struct vm_area_struct *, unsigned long, unsigned long, struct page *); void __unmap_hugepage_range_final(struct mmu_gather *tlb, diff --git a/include/linux/mm.h b/include/linux/mm.h index 322ec61d0da7..1dcd1ac81992 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1824,7 +1824,8 @@ long get_user_pages_locked(unsigned long start, unsigned long nr_pages, long pin_user_pages_locked(unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, int *locked); long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages, - struct page **pages, unsigned int gup_flags); + struct page **pages, unsigned int gup_flags, + int *fault_error); long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages, struct page **pages, unsigned int gup_flags); diff --git a/mm/gup.c b/mm/gup.c index 0697134b6a12..ab55a67aef78 100644 --- a/mm/gup.c +++ b/mm/gup.c @@ -881,7 +881,8 @@ static int get_gate_page(struct mm_struct *mm, unsigned long address, * is, *@locked will be set to 0 and -EBUSY returned. */ static int faultin_page(struct vm_area_struct *vma, - unsigned long address, unsigned int *flags, int *locked) + unsigned long address, unsigned int *flags, int *locked, + int *fault_error) { unsigned int fault_flags = 0; vm_fault_t ret; @@ -906,6 +907,8 @@ static int faultin_page(struct vm_area_struct *vma, } ret = handle_mm_fault(vma, address, fault_flags, NULL); + if (fault_error) + *fault_error = ret; if (ret & VM_FAULT_ERROR) { int err = vm_fault_to_errno(ret, *flags); @@ -996,6 +999,8 @@ static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags) * @vmas: array of pointers to vmas corresponding to each page. * Or NULL if the caller does not require them. * @locked: whether we're still with the mmap_lock held + * @fault_error: VM fault error from handle_mm_fault. NULL if the caller + * does not require this error. * * Returns either number of pages pinned (which may be less than the * number requested), or an error. Details about the return value: @@ -1040,6 +1045,13 @@ static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags) * when it's been released. Otherwise, it must be held for either * reading or writing and will not be released. * + * If @fault_error != NULL, __get_user_pages will return the VM fault error + * from handle_mm_fault() in this argument in the event of a VM fault error. + * On success (ret == nr_pages) fault_error is zero. + * On failure (ret != nr_pages) fault_error may still be 0 if the error did + * not originate from handle_mm_fault(). + * + * * In most cases, get_user_pages or get_user_pages_fast should be used * instead of __get_user_pages. __get_user_pages should be used only if * you need some special @gup_flags. @@ -1047,7 +1059,8 @@ static int check_vma_flags(struct vm_area_struct *vma, unsigned long gup_flags) static long __get_user_pages(struct mm_struct *mm, unsigned long start, unsigned long nr_pages, unsigned int gup_flags, struct page **pages, - struct vm_area_struct **vmas, int *locked) + struct vm_area_struct **vmas, int *locked, + int *fault_error) { long ret = 0, i = 0; struct vm_area_struct *vma = NULL; @@ -1097,7 +1110,7 @@ static long __get_user_pages(struct mm_struct *mm, if (is_vm_hugetlb_page(vma)) { i = follow_hugetlb_page(mm, vma, pages, vmas, &start, &nr_pages, i, - gup_flags, locked); + gup_flags, locked, fault_error); if (locked && *locked == 0) { /* * We've got a VM_FAULT_RETRY @@ -1124,7 +1137,8 @@ static long __get_user_pages(struct mm_struct *mm, page = follow_page_mask(vma, start, foll_flags, &ctx); if (!page) { - ret = faultin_page(vma, start, &foll_flags, locked); + ret = faultin_page(vma, start, &foll_flags, locked, + fault_error); switch (ret) { case 0: goto retry; @@ -1280,7 +1294,8 @@ static __always_inline long __get_user_pages_locked(struct mm_struct *mm, struct page **pages, struct vm_area_struct **vmas, int *locked, - unsigned int flags) + unsigned int flags, + int *fault_error) { long ret, pages_done; bool lock_dropped; @@ -1311,7 +1326,7 @@ static __always_inline long __get_user_pages_locked(struct mm_struct *mm, lock_dropped = false; for (;;) { ret = __get_user_pages(mm, start, nr_pages, flags, pages, - vmas, locked); + vmas, locked, fault_error); if (!locked) /* VM_FAULT_RETRY couldn't trigger, bypass */ return ret; @@ -1371,7 +1386,7 @@ static __always_inline long __get_user_pages_locked(struct mm_struct *mm, *locked = 1; ret = __get_user_pages(mm, start, 1, flags | FOLL_TRIED, - pages, NULL, locked); + pages, NULL, locked, fault_error); if (!*locked) { /* Continue to retry until we succeeded */ BUG_ON(ret != 0); @@ -1458,7 +1473,7 @@ long populate_vma_page_range(struct vm_area_struct *vma, * not result in a stack expansion that recurses back here. */ return __get_user_pages(mm, start, nr_pages, gup_flags, - NULL, NULL, locked); + NULL, NULL, locked, NULL); } /* @@ -1524,7 +1539,7 @@ int __mm_populate(unsigned long start, unsigned long len, int ignore_errors) static long __get_user_pages_locked(struct mm_struct *mm, unsigned long start, unsigned long nr_pages, struct page **pages, struct vm_area_struct **vmas, int *locked, - unsigned int foll_flags) + unsigned int foll_flags, int *fault_error) { struct vm_area_struct *vma; unsigned long vm_flags; @@ -1590,7 +1605,8 @@ struct page *get_dump_page(unsigned long addr) if (mmap_read_lock_killable(mm)) return NULL; ret = __get_user_pages_locked(mm, addr, 1, &page, NULL, &locked, - FOLL_FORCE | FOLL_DUMP | FOLL_GET); + FOLL_FORCE | FOLL_DUMP | FOLL_GET, + NULL); if (locked) mmap_read_unlock(mm); @@ -1704,11 +1720,11 @@ static long __gup_longterm_locked(struct mm_struct *mm, if (!(gup_flags & FOLL_LONGTERM)) return __get_user_pages_locked(mm, start, nr_pages, pages, vmas, - NULL, gup_flags); + NULL, gup_flags, NULL); flags = memalloc_pin_save(); do { rc = __get_user_pages_locked(mm, start, nr_pages, pages, vmas, - NULL, gup_flags); + NULL, gup_flags, NULL); if (rc <= 0) break; rc = check_and_migrate_movable_pages(rc, pages, gup_flags); @@ -1764,7 +1780,8 @@ static long __get_user_pages_remote(struct mm_struct *mm, return __get_user_pages_locked(mm, start, nr_pages, pages, vmas, locked, - gup_flags | FOLL_TOUCH | FOLL_REMOTE); + gup_flags | FOLL_TOUCH | FOLL_REMOTE, + NULL); } /** @@ -1941,7 +1958,7 @@ long get_user_pages_locked(unsigned long start, unsigned long nr_pages, return __get_user_pages_locked(current->mm, start, nr_pages, pages, NULL, locked, - gup_flags | FOLL_TOUCH); + gup_flags | FOLL_TOUCH, NULL); } EXPORT_SYMBOL(get_user_pages_locked); @@ -1961,7 +1978,8 @@ EXPORT_SYMBOL(get_user_pages_locked); * (e.g. FOLL_FORCE) are not required. */ long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages, - struct page **pages, unsigned int gup_flags) + struct page **pages, unsigned int gup_flags, + int *fault_error) { struct mm_struct *mm = current->mm; int locked = 1; @@ -1978,7 +1996,8 @@ long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages, mmap_read_lock(mm); ret = __get_user_pages_locked(mm, start, nr_pages, pages, NULL, - &locked, gup_flags | FOLL_TOUCH); + &locked, gup_flags | FOLL_TOUCH, + fault_error); if (locked) mmap_read_unlock(mm); return ret; @@ -2550,7 +2569,7 @@ static int __gup_longterm_unlocked(unsigned long start, int nr_pages, mmap_read_unlock(current->mm); } else { ret = get_user_pages_unlocked(start, nr_pages, - pages, gup_flags); + pages, gup_flags, NULL); } return ret; @@ -2880,7 +2899,7 @@ long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages, return -EINVAL; gup_flags |= FOLL_PIN; - return get_user_pages_unlocked(start, nr_pages, pages, gup_flags); + return get_user_pages_unlocked(start, nr_pages, pages, gup_flags, NULL); } EXPORT_SYMBOL(pin_user_pages_unlocked); @@ -2909,6 +2928,6 @@ long pin_user_pages_locked(unsigned long start, unsigned long nr_pages, gup_flags |= FOLL_PIN; return __get_user_pages_locked(current->mm, start, nr_pages, pages, NULL, locked, - gup_flags | FOLL_TOUCH); + gup_flags | FOLL_TOUCH, NULL); } EXPORT_SYMBOL(pin_user_pages_locked); diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 3db405dea3dc..889ac33d57d5 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -5017,7 +5017,8 @@ static void record_subpages_vmas(struct page *page, struct vm_area_struct *vma, long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma, struct page **pages, struct vm_area_struct **vmas, unsigned long *position, unsigned long *nr_pages, - long i, unsigned int flags, int *locked) + long i, unsigned int flags, int *locked, + int *fault_error) { unsigned long pfn_offset; unsigned long vaddr = *position; @@ -5103,6 +5104,8 @@ long follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma, } ret = hugetlb_fault(mm, vma, vaddr, fault_flags); if (ret & VM_FAULT_ERROR) { + if (fault_error) + *fault_error = ret; err = vm_fault_to_errno(ret, flags); remainder = 0; break; diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c index 2799c6660cce..0a20d926ae32 100644 --- a/virt/kvm/kvm_main.c +++ b/virt/kvm/kvm_main.c @@ -2004,6 +2004,30 @@ static bool hva_to_pfn_fast(unsigned long addr, bool write_fault, return false; } +static void kvm_send_vm_fault_signal(int fault_error, int errno, + unsigned long address, + struct task_struct *tsk) +{ + kernel_siginfo_t info; + + clear_siginfo(&info); + + if (fault_error == VM_FAULT_SIGBUS) + info.si_signo = SIGBUS; + else if (fault_error == VM_FAULT_SIGSEGV) + info.si_signo = SIGSEGV; + else + // Other fault errors should not result in a signal. + return; + + info.si_errno = errno; + info.si_code = BUS_ADRERR; + info.si_addr = (void __user *)address; + info.si_addr_lsb = PAGE_SHIFT; + + send_sig_info(info.si_signo, &info, tsk); +} + /* * The slow path to get the pfn of the specified host virtual address, * 1 indicates success, -errno is returned if error is detected. @@ -2014,6 +2038,7 @@ static int hva_to_pfn_slow(unsigned long addr, bool *async, bool write_fault, unsigned int flags = FOLL_HWPOISON; struct page *page; int npages = 0; + int fault_error; might_sleep(); @@ -2025,7 +2050,10 @@ static int hva_to_pfn_slow(unsigned long addr, bool *async, bool write_fault, if (async) flags |= FOLL_NOWAIT; - npages = get_user_pages_unlocked(addr, 1, &page, flags); + npages = get_user_pages_unlocked(addr, 1, &page, flags, &fault_error); + if (fault_error & VM_FAULT_ERROR) + kvm_send_vm_fault_signal(fault_error, npages, addr, current); + if (npages != 1) return npages; -- 2.31.1.751.gd2f1c929bd-goog

2 years, 11 months

3
5
0 0

[PATCH 1/2] kselftest/vm: Rename TARGETS to BUILD_TARGETS

by Alistair Popple

The TARGETS variable can be used when building selftests to specify a subset of selftests to build and run like so: make -C tools/testing/selftests TARGETS=ptrace run_tests However specifying TARGETS=vm results in the following build error as the TARGETS variable is used internally in the Makefile: make[1]: *** No rule to make target 'vm.c', needed by 'linux/kselftest/vm/vm_32'. Stop. Fix this by renaming the vm Makefile TARGETS variable to BUILD_TARGETS. Signed-off-by: Alistair Popple <apopple(a)nvidia.com> --- tools/testing/selftests/vm/Makefile | 10 +++++----- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/vm/Makefile b/tools/testing/selftests/vm/Makefile index 73e1cc96d7c2..110751ce8701 100644 --- a/tools/testing/selftests/vm/Makefile +++ b/tools/testing/selftests/vm/Makefile @@ -49,9 +49,9 @@ CAN_BUILD_I386 := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_32bit_progra CAN_BUILD_X86_64 := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_64bit_program.c) CAN_BUILD_WITH_NOPIE := $(shell ./../x86/check_cc.sh $(CC) ../x86/trivial_program.c -no-pie) -TARGETS := protection_keys -BINARIES_32 := $(TARGETS:%=%_32) -BINARIES_64 := $(TARGETS:%=%_64) +BUILD_TARGETS := protection_keys +BINARIES_32 := $(BUILD_TARGETS:%=%_32) +BINARIES_64 := $(BUILD_TARGETS:%=%_64) ifeq ($(CAN_BUILD_WITH_NOPIE),1) CFLAGS += -no-pie @@ -104,7 +104,7 @@ $(BINARIES_32): CFLAGS += -m32 $(BINARIES_32): LDLIBS += -lrt -ldl -lm $(BINARIES_32): $(OUTPUT)/%_32: %.c $(CC) $(CFLAGS) $(EXTRA_CFLAGS) $(notdir $^) $(LDLIBS) -o $@ -$(foreach t,$(TARGETS),$(eval $(call gen-target-rule-32,$(t)))) +$(foreach t,$(BUILD_TARGETS),$(eval $(call gen-target-rule-32,$(t)))) endif ifeq ($(CAN_BUILD_X86_64),1) @@ -112,7 +112,7 @@ $(BINARIES_64): CFLAGS += -m64 $(BINARIES_64): LDLIBS += -lrt -ldl $(BINARIES_64): $(OUTPUT)/%_64: %.c $(CC) $(CFLAGS) $(EXTRA_CFLAGS) $(notdir $^) $(LDLIBS) -o $@ -$(foreach t,$(TARGETS),$(eval $(call gen-target-rule-64,$(t)))) +$(foreach t,$(BUILD_TARGETS),$(eval $(call gen-target-rule-64,$(t)))) endif # x86_64 users should be encouraged to install 32-bit libraries -- 2.20.1

2 years, 11 months

1
1
0 0

[PATCH v2 1/2] KVM: selftests: Keep track of memslots more efficiently

by Maciej S. Szmigiero

From: "Maciej S. Szmigiero" <maciej.szmigiero(a)oracle.com> The KVM selftest framework was using a simple list for keeping track of the memslots currently in use. This resulted in lookups and adding a single memslot being O(n), the later due to linear scanning of the existing memslot set to check for the presence of any conflicting entries. Before this change, benchmarking high count of memslots was more or less impossible as pretty much all the benchmark time was spent in the selftest framework code. We can simply use a rbtree for keeping track of both of gfn and hva. We don't need an interval tree for hva here as we can't have overlapping memslots because we allocate a completely new memory chunk for each new memslot. Signed-off-by: Maciej S. Szmigiero <maciej.szmigiero(a)oracle.com> Reviewed-by: Andrew Jones <drjones(a)redhat.com> --- Changes from v1: * Add Andrew's Reviewed-by: tag tools/testing/selftests/kvm/Makefile | 2 +- tools/testing/selftests/kvm/lib/kvm_util.c | 141 ++++++++++++++---- .../selftests/kvm/lib/kvm_util_internal.h | 15 +- tools/testing/selftests/kvm/lib/rbtree.c | 1 + 4 files changed, 124 insertions(+), 35 deletions(-) create mode 100644 tools/testing/selftests/kvm/lib/rbtree.c diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile index a6d61f451f88..c30a21c1d676 100644 --- a/tools/testing/selftests/kvm/Makefile +++ b/tools/testing/selftests/kvm/Makefile @@ -33,7 +33,7 @@ ifeq ($(ARCH),s390) UNAME_M := s390x endif -LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c lib/test_util.c lib/guest_modes.c lib/perf_test_util.c +LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/rbtree.c lib/sparsebit.c lib/test_util.c lib/guest_modes.c lib/perf_test_util.c LIBKVM_x86_64 = lib/x86_64/processor.c lib/x86_64/vmx.c lib/x86_64/svm.c lib/x86_64/ucall.c lib/x86_64/handlers.S LIBKVM_aarch64 = lib/aarch64/processor.c lib/aarch64/ucall.c LIBKVM_s390x = lib/s390x/processor.c lib/s390x/ucall.c lib/s390x/diag318_test_handler.c diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c index e5fbf16f725b..69ee0a72c7d8 100644 --- a/tools/testing/selftests/kvm/lib/kvm_util.c +++ b/tools/testing/selftests/kvm/lib/kvm_util.c @@ -197,7 +197,9 @@ struct kvm_vm *vm_create(enum vm_guest_mode mode, uint64_t phy_pages, int perm) TEST_ASSERT(vm != NULL, "Insufficient Memory"); INIT_LIST_HEAD(&vm->vcpus); - INIT_LIST_HEAD(&vm->userspace_mem_regions); + vm->regions.gpa_tree = RB_ROOT; + vm->regions.hva_tree = RB_ROOT; + hash_init(vm->regions.slot_hash); vm->mode = mode; vm->type = 0; @@ -349,13 +351,14 @@ struct kvm_vm *vm_create_default(uint32_t vcpuid, uint64_t extra_mem_pages, */ void kvm_vm_restart(struct kvm_vm *vmp, int perm) { + int ctr; struct userspace_mem_region *region; vm_open(vmp, perm); if (vmp->has_irqchip) vm_create_irqchip(vmp); - list_for_each_entry(region, &vmp->userspace_mem_regions, list) { + hash_for_each(vmp->regions.slot_hash, ctr, region, slot_node) { int ret = ioctl(vmp->fd, KVM_SET_USER_MEMORY_REGION, &region->region); TEST_ASSERT(ret == 0, "KVM_SET_USER_MEMORY_REGION IOCTL failed,\n" " rc: %i errno: %i\n" @@ -418,14 +421,21 @@ uint32_t kvm_vm_reset_dirty_ring(struct kvm_vm *vm) static struct userspace_mem_region * userspace_mem_region_find(struct kvm_vm *vm, uint64_t start, uint64_t end) { - struct userspace_mem_region *region; + struct rb_node *node; - list_for_each_entry(region, &vm->userspace_mem_regions, list) { + for (node = vm->regions.gpa_tree.rb_node; node; ) { + struct userspace_mem_region *region = + container_of(node, struct userspace_mem_region, gpa_node); uint64_t existing_start = region->region.guest_phys_addr; uint64_t existing_end = region->region.guest_phys_addr + region->region.memory_size - 1; if (start <= existing_end && end >= existing_start) return region; + + if (start < existing_start) + node = node->rb_left; + else + node = node->rb_right; } return NULL; @@ -540,11 +550,16 @@ void kvm_vm_release(struct kvm_vm *vmp) } static void __vm_mem_region_delete(struct kvm_vm *vm, - struct userspace_mem_region *region) + struct userspace_mem_region *region, + bool unlink) { int ret; - list_del(&region->list); + if (unlink) { + rb_erase(&region->gpa_node, &vm->regions.gpa_tree); + rb_erase(&region->hva_node, &vm->regions.hva_tree); + hash_del(&region->slot_node); + } region->region.memory_size = 0; ret = ioctl(vm->fd, KVM_SET_USER_MEMORY_REGION, &region->region); @@ -563,14 +578,16 @@ static void __vm_mem_region_delete(struct kvm_vm *vm, */ void kvm_vm_free(struct kvm_vm *vmp) { - struct userspace_mem_region *region, *tmp; + int ctr; + struct hlist_node *node; + struct userspace_mem_region *region; if (vmp == NULL) return; /* Free userspace_mem_regions. */ - list_for_each_entry_safe(region, tmp, &vmp->userspace_mem_regions, list) - __vm_mem_region_delete(vmp, region); + hash_for_each_safe(vmp->regions.slot_hash, ctr, node, region, slot_node) + __vm_mem_region_delete(vmp, region, false); /* Free sparsebit arrays. */ sparsebit_free(&vmp->vpages_valid); @@ -652,6 +669,57 @@ int kvm_memcmp_hva_gva(void *hva, struct kvm_vm *vm, vm_vaddr_t gva, size_t len) return 0; } +static void vm_userspace_mem_region_gpa_insert(struct rb_root *gpa_tree, + struct userspace_mem_region *region) +{ + struct rb_node **cur, *parent; + + for (cur = &gpa_tree->rb_node, parent = NULL; *cur; ) { + struct userspace_mem_region *cregion; + + cregion = container_of(*cur, typeof(*cregion), gpa_node); + parent = *cur; + if (region->region.guest_phys_addr < + cregion->region.guest_phys_addr) + cur = &(*cur)->rb_left; + else { + TEST_ASSERT(region->region.guest_phys_addr != + cregion->region.guest_phys_addr, + "Duplicate GPA in region tree"); + + cur = &(*cur)->rb_right; + } + } + + rb_link_node(&region->gpa_node, parent, cur); + rb_insert_color(&region->gpa_node, gpa_tree); +} + +static void vm_userspace_mem_region_hva_insert(struct rb_root *hva_tree, + struct userspace_mem_region *region) +{ + struct rb_node **cur, *parent; + + for (cur = &hva_tree->rb_node, parent = NULL; *cur; ) { + struct userspace_mem_region *cregion; + + cregion = container_of(*cur, typeof(*cregion), hva_node); + parent = *cur; + if (region->host_mem < cregion->host_mem) + cur = &(*cur)->rb_left; + else { + TEST_ASSERT(region->host_mem != + cregion->host_mem, + "Duplicate HVA in region tree"); + + cur = &(*cur)->rb_right; + } + } + + rb_link_node(&region->hva_node, parent, cur); + rb_insert_color(&region->hva_node, hva_tree); +} + /* * VM Userspace Memory Region Add * @@ -716,7 +784,8 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, (uint64_t) region->region.memory_size); /* Confirm no region with the requested slot already exists. */ - list_for_each_entry(region, &vm->userspace_mem_regions, list) { + hash_for_each_possible(vm->regions.slot_hash, region, slot_node, + slot) { if (region->region.slot != slot) continue; @@ -796,8 +865,10 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm, ret, errno, slot, flags, guest_paddr, (uint64_t) region->region.memory_size); - /* Add to linked-list of memory regions. */ - list_add(&region->list, &vm->userspace_mem_regions); + /* Add to quick lookup data structures */ + vm_userspace_mem_region_gpa_insert(&vm->regions.gpa_tree, region); + vm_userspace_mem_region_hva_insert(&vm->regions.hva_tree, region); + hash_add(vm->regions.slot_hash, &region->slot_node, slot); } /* @@ -820,10 +891,10 @@ memslot2region(struct kvm_vm *vm, uint32_t memslot) { struct userspace_mem_region *region; - list_for_each_entry(region, &vm->userspace_mem_regions, list) { + hash_for_each_possible(vm->regions.slot_hash, region, slot_node, + memslot) if (region->region.slot == memslot) return region; - } fprintf(stderr, "No mem region with the requested slot found,\n" " requested slot: %u\n", memslot); @@ -908,7 +979,7 @@ void vm_mem_region_move(struct kvm_vm *vm, uint32_t slot, uint64_t new_gpa) */ void vm_mem_region_delete(struct kvm_vm *vm, uint32_t slot) { - __vm_mem_region_delete(vm, memslot2region(vm, slot)); + __vm_mem_region_delete(vm, memslot2region(vm, slot), true); } /* @@ -1180,16 +1251,14 @@ void *addr_gpa2hva(struct kvm_vm *vm, vm_paddr_t gpa) { struct userspace_mem_region *region; - list_for_each_entry(region, &vm->userspace_mem_regions, list) { - if ((gpa >= region->region.guest_phys_addr) - && (gpa <= (region->region.guest_phys_addr - + region->region.memory_size - 1))) - return (void *) ((uintptr_t) region->host_mem - + (gpa - region->region.guest_phys_addr)); + region = userspace_mem_region_find(vm, gpa, gpa); + if (!region) { + TEST_FAIL("No vm physical memory at 0x%lx", gpa); + return NULL; } - TEST_FAIL("No vm physical memory at 0x%lx", gpa); - return NULL; + return (void *)((uintptr_t)region->host_mem + + (gpa - region->region.guest_phys_addr)); } /* @@ -1211,15 +1280,22 @@ void *addr_gpa2hva(struct kvm_vm *vm, vm_paddr_t gpa) */ vm_paddr_t addr_hva2gpa(struct kvm_vm *vm, void *hva) { - struct userspace_mem_region *region; + struct rb_node *node; + + for (node = vm->regions.hva_tree.rb_node; node; ) { + struct userspace_mem_region *region = + container_of(node, struct userspace_mem_region, hva_node); + + if (hva >= region->host_mem) { + if (hva <= (region->host_mem + + region->region.memory_size - 1)) + return (vm_paddr_t)((uintptr_t) + region->region.guest_phys_addr + + (hva - (uintptr_t)region->host_mem)); - list_for_each_entry(region, &vm->userspace_mem_regions, list) { - if ((hva >= region->host_mem) - && (hva <= (region->host_mem - + region->region.memory_size - 1))) - return (vm_paddr_t) ((uintptr_t) - region->region.guest_phys_addr - + (hva - (uintptr_t) region->host_mem)); + node = node->rb_right; + } else + node = node->rb_left; } TEST_FAIL("No mapping to a guest physical address, hva: %p", hva); @@ -1745,6 +1821,7 @@ int _kvm_ioctl(struct kvm_vm *vm, unsigned long cmd, void *arg) */ void vm_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent) { + int ctr; struct userspace_mem_region *region; struct vcpu *vcpu; @@ -1752,7 +1829,7 @@ void vm_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent) fprintf(stream, "%*sfd: %i\n", indent, "", vm->fd); fprintf(stream, "%*spage_size: 0x%x\n", indent, "", vm->page_size); fprintf(stream, "%*sMem Regions:\n", indent, ""); - list_for_each_entry(region, &vm->userspace_mem_regions, list) { + hash_for_each(vm->regions.slot_hash, ctr, region, slot_node) { fprintf(stream, "%*sguest_phys: 0x%lx size: 0x%lx " "host_virt: %p\n", indent + 2, "", (uint64_t) region->region.guest_phys_addr, diff --git a/tools/testing/selftests/kvm/lib/kvm_util_internal.h b/tools/testing/selftests/kvm/lib/kvm_util_internal.h index 34465dc562d8..af310110602b 100644 --- a/tools/testing/selftests/kvm/lib/kvm_util_internal.h +++ b/tools/testing/selftests/kvm/lib/kvm_util_internal.h @@ -8,6 +8,9 @@ #ifndef SELFTEST_KVM_UTIL_INTERNAL_H #define SELFTEST_KVM_UTIL_INTERNAL_H +#include "linux/hashtable.h" +#include "linux/rbtree.h" + #include "sparsebit.h" #define KVM_DEV_PATH "/dev/kvm" @@ -20,7 +23,9 @@ struct userspace_mem_region { void *host_mem; void *mmap_start; size_t mmap_size; - struct list_head list; + struct rb_node gpa_node; + struct rb_node hva_node; + struct hlist_node slot_node; }; struct vcpu { @@ -33,6 +38,12 @@ struct vcpu { uint32_t dirty_gfns_count; }; +struct userspace_mem_regions { + struct rb_root gpa_tree; + struct rb_root hva_tree; + DECLARE_HASHTABLE(slot_hash, 9); +}; + struct kvm_vm { int mode; unsigned long type; @@ -45,7 +56,7 @@ struct kvm_vm { unsigned int va_bits; uint64_t max_gfn; struct list_head vcpus; - struct list_head userspace_mem_regions; + struct userspace_mem_regions regions; struct sparsebit *vpages_valid; struct sparsebit *vpages_mapped; bool has_irqchip; diff --git a/tools/testing/selftests/kvm/lib/rbtree.c b/tools/testing/selftests/kvm/lib/rbtree.c new file mode 100644 index 000000000000..a703f0194ea3 --- /dev/null +++ b/tools/testing/selftests/kvm/lib/rbtree.c @@ -0,0 +1 @@ +#include "../../../../lib/rbtree.c"

2 years, 11 months

2
2
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror May 2021