August 2023 - Linux-kselftest-mirror

[PATCH v4 bpf 0/2] bpf: return proper error codes for lwt redirect

by Yan Zhai

lwt xmit hook does not expect positive return values in function ip_finish_output2 and ip6_finish_output2. However, BPF redirect programs can return positive values such like NET_XMIT_DROP, NET_RX_DROP, and etc as errors. Such return values can panic the kernel unexpectedly: https://gist.github.com/zhaiyan920/8fbac245b261fe316a7ef04c9b1eba48 This patch fixes the return values from BPF redirect, so the error handling would be consistent at xmit hook. It also adds a few test cases to prevent future regressions. v3: https://lore.kernel.org/bpf/cover.1690255889.git.yan@cloudflare.com/ v2: https://lore.kernel.org/netdev/ZLdY6JkWRccunvu0@debian.debian/ v1: https://lore.kernel.org/bpf/ZLbYdpWC8zt9EJtq@debian.debian/ changes since v3: * minor change in commit message and changelogs * tested by Jakub Sitnicki changes since v2: * subject name changed * also covered redirect to ingress case * added selftests changes since v1: * minor code style changes Yan Zhai (2): bpf: fix skb_do_redirect return values bpf: selftests: add lwt redirect regression test cases include/linux/netdevice.h | 2 + net/core/filter.c | 9 +- tools/testing/selftests/bpf/Makefile | 1 + .../selftests/bpf/progs/test_lwt_redirect.c | 66 +++++++ .../selftests/bpf/test_lwt_redirect.sh | 174 ++++++++++++++++++ 5 files changed, 250 insertions(+), 2 deletions(-) create mode 100644 tools/testing/selftests/bpf/progs/test_lwt_redirect.c create mode 100755 tools/testing/selftests/bpf/test_lwt_redirect.sh -- 2.30.2

1 year, 4 months

5
19
0 0

[PATCH v3 0/2] tools/nolibc: add pipe(), pipe2() and their testcase

by Yuan Tan

Hi Willy and Thomas, In v3, I have fixed all the problems you mentioned. Welcome any other suggestion. --- Changes in v3: - Fix the missing return - Fix __NR_pipe to __NR_pipe2 - Fix the missing static - Test case works in one process - Link to v2: https://lore.kernel.org/all/cover.1690733545.git.tanyuan@tinylab.org Changes in v2: - Use sys_pipe2 to implement the pipe() - Use memcmp() instead of strcmp() - Link to v1: https://lore.kernel.org/all/cover.1690307717.git.tanyuan@tinylab.org --- Yuan Tan (2): tools/nolibc: add pipe() and pipe2() support selftests/nolibc: add testcase for pipe tools/include/nolibc/sys.h | 24 ++++++++++++++++++++ tools/testing/selftests/nolibc/nolibc-test.c | 22 ++++++++++++++++++ 2 files changed, 46 insertions(+) -- 2.34.1

1 year, 4 months

3
4
0 0

[PATCH v1 1/1] test_firmware: fix a memory leak with reqs buffer

by Mirsad Todorovac

[ commit be37bed754ed90b2655382f93f9724b3c1aae847 upstream ] Dan Carpenter spotted that test_fw_config->reqs will be leaked if trigger_batched_requests_store() is called two or more times. The same appears with trigger_batched_requests_async_store(). This bug wasn't triggered by the tests, but observed by Dan's visual inspection of the code. The recommended workaround was to return -EBUSY if test_fw_config->reqs is already allocated. Fixes: c92316bf8e94 ("test_firmware: add batched firmware tests") Cc: Luis Chamberlain <mcgrof(a)kernel.org> Cc: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Cc: Russ Weight <russell.h.weight(a)intel.com> Cc: Tianfei Zhang <tianfei.zhang(a)intel.com> Cc: Shuah Khan <shuah(a)kernel.org> Cc: Colin Ian King <colin.i.king(a)gmail.com> Cc: Randy Dunlap <rdunlap(a)infradead.org> Cc: linux-kselftest(a)vger.kernel.org Cc: stable(a)vger.kernel.org # v4.19 Suggested-by: Dan Carpenter <error27(a)gmail.com> Suggested-by: Takashi Iwai <tiwai(a)suse.de> Reviewed-by: Dan Carpenter <dan.carpenter(a)linaro.org> Acked-by: Luis Chamberlain <mcgrof(a)kernel.org> Link: https://lore.kernel.org/r/20230509084746.48259-2-mirsad.todorovac@alu.unizg… Signed-off-by: Mirsad Todorovac <mirsad.todorovac(a)alu.unizg.hr> [ This is a backport to v4.19 stable branch without a change in code from the 5.4+ patch ] --- v1: patch sumbmitted verbatim from the 5.4+ branch to 4.19 lib/test_firmware.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/lib/test_firmware.c b/lib/test_firmware.c index f4cc874021da..e4688821eab8 100644 --- a/lib/test_firmware.c +++ b/lib/test_firmware.c @@ -618,6 +618,11 @@ static ssize_t trigger_batched_requests_store(struct device *dev, mutex_lock(&test_fw_mutex); + if (test_fw_config->reqs) { + rc = -EBUSY; + goto out_bail; + } + test_fw_config->reqs = vzalloc(array3_size(sizeof(struct test_batched_req), test_fw_config->num_requests, 2)); @@ -721,6 +726,11 @@ ssize_t trigger_batched_requests_async_store(struct device *dev, mutex_lock(&test_fw_mutex); + if (test_fw_config->reqs) { + rc = -EBUSY; + goto out_bail; + } + test_fw_config->reqs = vzalloc(array3_size(sizeof(struct test_batched_req), test_fw_config->num_requests, 2)); -- 2.34.1

1 year, 4 months

1
0
0 0

[PATCH v2 0/2] tools/nolibc: add pipe(), pipe2() and their testcase

by Yuan Tan

Hi, Willy, Thomas, Thanks to your advice and I really learned a lot from it. V2 now uses pipe2() to wrap pipe(), and fixes the strcmp issue in test case. Best regards, Yuan Tan Yuan Tan (2): tools/nolibc: add pipe() and pipe2() support selftests/nolibc: add testcase for pipe tools/include/nolibc/sys.h | 24 ++++++++++++++ tools/testing/selftests/nolibc/nolibc-test.c | 35 ++++++++++++++++++++ 2 files changed, 59 insertions(+) -- 2.34.1

1 year, 4 months

2
15
0 0

[PATCH 0/3] rust: kunit: Support KUnit tests with a user-space like syntax

by David Gow

This series was originally written by José Expósito, and can be found here: https://github.com/Rust-for-Linux/linux/pull/950 Add support for writing KUnit tests in Rust. While Rust doctests are already converted to KUnit tests and run, they're really better suited for examples, rather than as first-class unit tests. This series implements a series of direct Rust bindings for KUnit tests, as well as a new macro which allows KUnit tests to be written using a close variant of normal Rust unit test syntax. The only change required is replacing '#[cfg(test)]' with '#[kunit_tests(kunit_test_suite_name)]' An example test would look like: #[kunit_tests(rust_kernel_hid_driver)] mod tests { use super::*; use crate::{c_str, driver, hid, prelude::*}; use core::ptr; struct SimpleTestDriver; impl Driver for SimpleTestDriver { type Data = (); } #[test] fn rust_test_hid_driver_adapter() { let mut hid = bindings::hid_driver::default(); let name = c_str!("SimpleTestDriver"); static MODULE: ThisModule = unsafe { ThisModule::from_ptr(ptr::null_mut()) }; let res = unsafe { <hid::Adapter<SimpleTestDriver> as driver::DriverOps>::register(&mut hid, name, &MODULE) }; assert_eq!(res, Err(ENODEV)); // The mock returns -19 } } Changes since the GitHub PR: - Rebased on top of kselftest/kunit - Add const_mut_refs feature This may conflict with https://lore.kernel.org/lkml/20230503090708.2524310-6-nmi@metaspace.dk/ - Add rust/macros/kunit.rs to the KUnit MAINTAINERS entry Signed-off-by: David Gow <davidgow(a)google.com> --- José Expósito (3): rust: kunit: add KUnit case and suite macros rust: macros: add macro to easily run KUnit tests rust: kunit: allow to know if we are in a test MAINTAINERS | 1 + rust/kernel/kunit.rs | 181 +++++++++++++++++++++++++++++++++++++++++++++++++++ rust/kernel/lib.rs | 1 + rust/macros/kunit.rs | 149 ++++++++++++++++++++++++++++++++++++++++++ rust/macros/lib.rs | 29 +++++++++ 5 files changed, 361 insertions(+) --- base-commit: 64bd4641310c41a1ecf07c13c67bc0ed61045dfd change-id: 20230720-rustbind-477964954da5 Best regards, -- David Gow <davidgow(a)google.com>

1 year, 4 months

3
7
0 0

[PATCH v1 1/1] test_firmware: fix the memory leaks with the reqs buffer

by Mirsad Todorovac

[ commit be37bed754ed90b2655382f93f9724b3c1aae847 upstream ] Dan Carpenter spotted that test_fw_config->reqs will be leaked if trigger_batched_requests_store() is called two or more times. The same appears with trigger_batched_requests_async_store(). This bug wasn't triggered by the tests, but observed by Dan's visual inspection of the code. The recommended workaround was to return -EBUSY if test_fw_config->reqs is already allocated. Fixes: c92316bf8e94 ("test_firmware: add batched firmware tests") Cc: Luis Chamberlain <mcgrof(a)kernel.org> Cc: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Cc: Russ Weight <russell.h.weight(a)intel.com> Cc: Tianfei Zhang <tianfei.zhang(a)intel.com> Cc: Shuah Khan <shuah(a)kernel.org> Cc: Colin Ian King <colin.i.king(a)gmail.com> Cc: Randy Dunlap <rdunlap(a)infradead.org> Cc: linux-kselftest(a)vger.kernel.org Cc: stable(a)vger.kernel.org # v4.14 Suggested-by: Dan Carpenter <error27(a)gmail.com> Suggested-by: Takashi Iwai <tiwai(a)suse.de> Reviewed-by: Dan Carpenter <dan.carpenter(a)linaro.org> Acked-by: Luis Chamberlain <mcgrof(a)kernel.org> Link: https://lore.kernel.org/r/20230509084746.48259-2-mirsad.todorovac@alu.unizg… Signed-off-by: Mirsad Todorovac <mirsad.todorovac(a)alu.unizg.hr> [ This fix is applied against the 4.14 stable branch. There are no changes to the ] [ fix in code when compared to the upstread, only the reformatting for backport. ] --- lib/test_firmware.c | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/lib/test_firmware.c b/lib/test_firmware.c index 1c5e5246bf10..5318c5e18acf 100644 --- a/lib/test_firmware.c +++ b/lib/test_firmware.c @@ -621,6 +621,11 @@ static ssize_t trigger_batched_requests_store(struct device *dev, mutex_lock(&test_fw_mutex); + if (test_fw_config->reqs) { + rc = -EBUSY; + goto out_bail; + } + test_fw_config->reqs = vzalloc(sizeof(struct test_batched_req) * test_fw_config->num_requests * 2); if (!test_fw_config->reqs) { @@ -723,6 +728,11 @@ ssize_t trigger_batched_requests_async_store(struct device *dev, mutex_lock(&test_fw_mutex); + if (test_fw_config->reqs) { + rc = -EBUSY; + goto out_bail; + } + test_fw_config->reqs = vzalloc(sizeof(struct test_batched_req) * test_fw_config->num_requests * 2); if (!test_fw_config->reqs) { -- 2.39.3

1 year, 4 months

1
0
0 0

tools/nolibc: RFC: report and drop unused functions/data by gc-section

by Zhangjin Wu

Hi, Willy, Thomas Currently, since this part of the discussion is out of the original topic [1], as suggested by Willy, open a new thread for this. [1]: https://lore.kernel.org/lkml/20230731224929.GA18296@1wt.eu/#R The original topic [1] tries to enable -Wall (or even -Wextra) to report some issues (include the dead unused functions/data) found during compiling stage, this further propose a method to enable '-ffunction-sections -fdata-sections -Wl,--gc-sections,--print-gc-sections to' find the dead unused functions/data during linking stage: Just thought about gc-section, do we need to further remove dead code/data in the binary? I don't think it is necessary for nolibc-test itself, but with '-Wl,--gc-sections -Wl,--print-gc-sections' may be a good helper to show us which ones should be dropped or which ones are wrongly declared as public? Just found '-O3 + -Wl,--gc-section + -Wl,--print-gc-sections' did tell us something as below: removing unused section '.text.nolibc_raise' removing unused section '.text.nolibc_memmove' removing unused section '.text.nolibc_abort' removing unused section '.text.nolibc_memcpy' removing unused section '.text.__stack_chk_init' removing unused section '.text.is_setting_valid' These info may help us further add missing 'static' keyword or find another method to to drop the wrongly used status of some functions from the code side. Let's continue the discussion as below. > On Mon, Jul 31, 2023 at 08:36:05PM +0200, Thomas Wei�schuh wrote: > > [...] > > > > > > It is very easy to add the missing 'static' keyword for is_setting_valid(), but > > > > > for __stack_chk_init(), is it ok for us to convert it to 'static' and remove > > > > > the 'weak' attrbute and even the 'section' attribute? seems it is only used by > > > > > our _start_c() currently. > > > > > > > > Making is_setting_valid(), __stack_chk_init() seems indeed useful. > > > > Also all the run_foo() test functions. > > > > > > Most of them could theoretically be turned to static. *But* it causes a > > > problem which is that it will multiply their occurrences in multi-unit > > > programs, and that's in part why we've started to use weak instead. Also > > > if you run through gdb and want to mark a break point, you won't have the > > > symbol when it's static, Willy, did I misunderstand something again? a simple test shows, seems this is not really always like that, static mainly means 'local', the symbol is still there if without -O2/-Os and is able to be set a breakpoint at: // test.c: gcc -o test test.c #include <stdio.h> static int test (void) { printf("hello, world!\n"); } int main(void) { test(); return 0; } Even with -Os/-O2, an additional '-g' is able to generate the 'test' symbol for debug as we expect. > and the code will appear at multiple locations, > > > which is really painful. I'd instead really prefer to avoid static when > > > we don't strictly want to inline the code, and prefer weak when possible > > > because we know many of them will be dropped at link time (and that's > > > the exact purpose). For the empty __stack_chk_init() one (when the arch not support stackprotector) we used, when with 'weak', it is not possible drop it during link time even with -O3, the weak one will be dropped may be only when there is a global one with the same name is used or the 'weak' one is never really used? #include <stdio.h> __attribute__((weak,unused,section(".text.nolibc_memset"))) int test (void) { printf("hello, world!\n"); } int main(void) { test(); return 0; } 0000000000001060 <main>: 1060: f3 0f 1e fa endbr64 1064: 48 83 ec 08 sub $0x8,%rsp 1068: e8 03 01 00 00 callq 1170 <test> 106d: 31 c0 xor %eax,%eax 106f: 48 83 c4 08 add $0x8,%rsp 1073: c3 retq 1074: 66 2e 0f 1f 84 00 00 nopw %cs:0x0(%rax,%rax,1) 107b: 00 00 00 107e: 66 90 xchg %ax,%ax Seems it is either impossible to add a 'inline' keyword again with the 'weak' attribute (warned by compiler), so, the _start_c (itself is always called by _start) will always add an empty call to the weak empty __stack_chk_init(), -Os/-O2/-O3 don't help. for such an empty function, in my opinion, as the size we want to care about, the calling place should be simply removed by compiler. Test also shows, with current __inline__ method, the calling place is removed, but with c89, the __stack_chk_init() itself will not be droped automatically, only if not with -std=c89, it will be dropped and not appear in the --print-gc-sections result. Even for a supported architecture, the shorter __stack_chk_init() may be better to inlined to the _start_c()? So, If my above test is ok, then, we'd better simply convert the whole __stack_chk_init() to a static one as below (I didn't investigate this deeply due to the warning about static and weak conflict at the first time): diff --git a/tools/include/nolibc/crt.h b/tools/include/nolibc/crt.h index 32e128b0fb62..a5f33fef1672 100644 --- a/tools/include/nolibc/crt.h +++ b/tools/include/nolibc/crt.h @@ -10,7 +10,7 @@ char **environ __attribute__((weak)); const unsigned long *_auxv __attribute__((weak)); -void __stack_chk_init(void); +static void __stack_chk_init(void); static void exit(int); void _start_c(long *sp) diff --git a/tools/include/nolibc/stackprotector.h b/tools/include/nolibc/stackprotector.h index b620f2b9578d..13f1d0e60387 100644 --- a/tools/include/nolibc/stackprotector.h +++ b/tools/include/nolibc/stackprotector.h @@ -37,8 +37,7 @@ void __stack_chk_fail_local(void) __attribute__((weak,section(".data.nolibc_stack_chk"))) uintptr_t __stack_chk_guard; -__attribute__((weak,section(".text.nolibc_stack_chk"))) __no_stack_protector -void __stack_chk_init(void) +static __no_stack_protector void __stack_chk_init(void) { my_syscall3(__NR_getrandom, &__stack_chk_guard, sizeof(__stack_chk_guard), 0); /* a bit more randomness in case getrandom() fails, ensure the guard is never 0 */ @@ -46,7 +45,7 @@ void __stack_chk_init(void) __stack_chk_guard ^= (uintptr_t) &__stack_chk_guard; } #else /* !defined(_NOLIBC_STACKPROTECTOR) */ -__inline__ void __stack_chk_init(void) {} +static void __stack_chk_init(void) {} #endif /* defined(_NOLIBC_STACKPROTECTOR) */ #endif /* _NOLIBC_STACKPROTECTOR_H */ > > > > Thanks for the clarification. I forgot about that completely! > > > > The stuff from nolibc-test.c itself (run_foo() and is_settings_valid()) > > should still be done. > > Yes, likely. Nolibc-test should be done just like users expect to use > nolibc, and nolibc should be the most flexible possible. For the 'static' keyword we tested above, '-g' may help the debug requirement, so, is ok for us to apply 'static' for them safely now? A further test shows, with 'static' on _start_c() doesn't help the size, for it is always called from _start(), will never save move instructions, but we need a more 'used' attribute to silence the error 'nolibc-test.c:(.text+0x38cd): undefined reference to `_start_c'', so, reserve it as the before simpler 'void _start_c(void)' may be better? static __attribute__((used)) void _start_c(long *sp) Thanks, Zhangjin > > Cheers, > Willy

1 year, 4 months

2
2
0 0

[PATCH] Modify the function to et the highest_perf value

by Meng Li

The previous function will be deprecated. Signed-off-by: Meng Li <li.meng(a)amd.com> --- drivers/cpufreq/amd-pstate-ut.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/cpufreq/amd-pstate-ut.c b/drivers/cpufreq/amd-pstate-ut.c index 7f3fe2048981..9c889a4a0177 100644 --- a/drivers/cpufreq/amd-pstate-ut.c +++ b/drivers/cpufreq/amd-pstate-ut.c @@ -145,8 +145,6 @@ static void amd_pstate_ut_check_perf(u32 index) struct cpufreq_policy *policy = NULL; struct amd_cpudata *cpudata = NULL; - highest_perf = amd_get_highest_perf(); - for_each_possible_cpu(cpu) { policy = cpufreq_cpu_get(cpu); if (!policy) @@ -161,6 +159,7 @@ static void amd_pstate_ut_check_perf(u32 index) return; } + highest_perf = cppc_perf.highest_perf; nominal_perf = cppc_perf.nominal_perf; lowest_nonlinear_perf = cppc_perf.lowest_nonlinear_perf; lowest_perf = cppc_perf.lowest_perf; @@ -172,6 +171,7 @@ static void amd_pstate_ut_check_perf(u32 index) return; } + highest_perf = AMD_CPPC_HIGHEST_PERF(cap1); nominal_perf = AMD_CPPC_NOMINAL_PERF(cap1); lowest_nonlinear_perf = AMD_CPPC_LOWNONLIN_PERF(cap1); lowest_perf = AMD_CPPC_LOWEST_PERF(cap1); -- 2.34.1

1 year, 4 months

3
2
0 0

[PATCH -next v2] selftests/bpf: replace fall through comment by fallthrough pseudo-keyword

by Ruan Jinjie

Replace the existing /* fall through */ comments with the new pseudo-keyword macro fallthrough[1]. [1] https://www.kernel.org/doc/html/v5.7/process/deprecated.html?highlight=fall… Signed-off-by: Ruan Jinjie <ruanjinjie(a)huawei.com> --- v2: - Update the subject and commit message. --- tools/testing/selftests/bpf/prog_tests/kfunc_call.c | 4 ++-- tools/testing/selftests/bpf/progs/test_cls_redirect.c | 2 +- tools/testing/selftests/bpf/progs/test_cls_redirect_dynptr.c | 2 +- tools/testing/selftests/bpf/test_verifier.c | 2 +- 4 files changed, 5 insertions(+), 5 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/kfunc_call.c b/tools/testing/selftests/bpf/prog_tests/kfunc_call.c index a543742cd7bd..0fd08172965a 100644 --- a/tools/testing/selftests/bpf/prog_tests/kfunc_call.c +++ b/tools/testing/selftests/bpf/prog_tests/kfunc_call.c @@ -101,7 +101,7 @@ static void verify_success(struct kfunc_test_params *param) case syscall_test: topts.ctx_in = &args; topts.ctx_size_in = sizeof(args); - /* fallthrough */ + fallthrough; case syscall_null_ctx_test: break; case tc_test: @@ -167,7 +167,7 @@ static void verify_fail(struct kfunc_test_params *param) case syscall_test: topts.ctx_in = &args; topts.ctx_size_in = sizeof(args); - /* fallthrough */ + fallthrough; case syscall_null_ctx_test: break; case tc_test: diff --git a/tools/testing/selftests/bpf/progs/test_cls_redirect.c b/tools/testing/selftests/bpf/progs/test_cls_redirect.c index 66b304982245..f97960759558 100644 --- a/tools/testing/selftests/bpf/progs/test_cls_redirect.c +++ b/tools/testing/selftests/bpf/progs/test_cls_redirect.c @@ -300,7 +300,7 @@ bool pkt_skip_ipv6_extension_headers(buf_t *pkt, case IPPROTO_FRAGMENT: *is_fragment = true; /* NB: We don't check that hdrlen == 0 as per spec. */ - /* fallthrough; */ + fallthrough; case IPPROTO_HOPOPTS: case IPPROTO_ROUTING: diff --git a/tools/testing/selftests/bpf/progs/test_cls_redirect_dynptr.c b/tools/testing/selftests/bpf/progs/test_cls_redirect_dynptr.c index f41c81212ee9..54dbf307c692 100644 --- a/tools/testing/selftests/bpf/progs/test_cls_redirect_dynptr.c +++ b/tools/testing/selftests/bpf/progs/test_cls_redirect_dynptr.c @@ -204,7 +204,7 @@ static bool pkt_skip_ipv6_extension_headers(struct bpf_dynptr *dynptr, __u64 *of case IPPROTO_FRAGMENT: *is_fragment = true; /* NB: We don't check that hdrlen == 0 as per spec. */ - /* fallthrough; */ + fallthrough; case IPPROTO_HOPOPTS: case IPPROTO_ROUTING: diff --git a/tools/testing/selftests/bpf/test_verifier.c b/tools/testing/selftests/bpf/test_verifier.c index 31f1c935cd07..5621a4e0a1be 100644 --- a/tools/testing/selftests/bpf/test_verifier.c +++ b/tools/testing/selftests/bpf/test_verifier.c @@ -1289,7 +1289,7 @@ static int do_prog_test_run(int fd_prog, bool unpriv, uint32_t expected_val, printf("Did not run the program (no permission) "); return 0; } - /* fallthrough; */ + fallthrough; default: printf("FAIL: Unexpected bpf_prog_test_run error (%s) ", strerror(saved_errno)); -- 2.34.1

1 year, 4 months

3
4
0 0

[PATCH v2 1/1] test_firmware: prevent race conditions by a correct implementation of locking

by Mirsad Todorovac

[ Upstream commit 4acfe3dfde685a5a9eaec5555351918e2d7266a1 ] Dan Carpenter spotted a race condition in a couple of situations like these in the test_firmware driver: static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) { u8 val; int ret; ret = kstrtou8(buf, 10, &val); if (ret) return ret; mutex_lock(&test_fw_mutex); *(u8 *)cfg = val; mutex_unlock(&test_fw_mutex); /* Always return full write size even if we didn't consume all */ return size; } static ssize_t config_num_requests_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { int rc; mutex_lock(&test_fw_mutex); if (test_fw_config->reqs) { pr_err("Must call release_all_firmware prior to changing config\n"); rc = -EINVAL; mutex_unlock(&test_fw_mutex); goto out; } mutex_unlock(&test_fw_mutex); // NOTE: HERE is the race!!! Function can be preempted! // test_fw_config->reqs can change between the release of // the lock about and acquire of the lock in the // test_dev_config_update_u8() rc = test_dev_config_update_u8(buf, count, &test_fw_config->num_requests); out: return rc; } static ssize_t config_read_fw_idx_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t count) { return test_dev_config_update_u8(buf, count, &test_fw_config->read_fw_idx); } The function test_dev_config_update_u8() is called from both the locked and the unlocked context, function config_num_requests_store() and config_read_fw_idx_store() which can both be called asynchronously as they are driver's methods, while test_dev_config_update_u8() and siblings change their argument pointed to by u8 *cfg or similar pointer. To avoid deadlock on test_fw_mutex, the lock is dropped before calling test_dev_config_update_u8() and re-acquired within test_dev_config_update_u8() itself, but alas this creates a race condition. Having two locks wouldn't assure a race-proof mutual exclusion. This situation is best avoided by the introduction of a new, unlocked function __test_dev_config_update_u8() which can be called from the locked context and reducing test_dev_config_update_u8() to: static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) { int ret; mutex_lock(&test_fw_mutex); ret = __test_dev_config_update_u8(buf, size, cfg); mutex_unlock(&test_fw_mutex); return ret; } doing the locking and calling the unlocked primitive, which enables both locked and unlocked versions without duplication of code. Fixes: c92316bf8e948 ("test_firmware: add batched firmware tests") Cc: Luis R. Rodriguez <mcgrof(a)kernel.org> Cc: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> Cc: Russ Weight <russell.h.weight(a)intel.com> Cc: Takashi Iwai <tiwai(a)suse.de> Cc: Tianfei Zhang <tianfei.zhang(a)intel.com> Cc: Shuah Khan <shuah(a)kernel.org> Cc: Colin Ian King <colin.i.king(a)gmail.com> Cc: Randy Dunlap <rdunlap(a)infradead.org> Cc: linux-kselftest(a)vger.kernel.org Cc: stable(a)vger.kernel.org # v5.4 Suggested-by: Dan Carpenter <error27(a)gmail.com> Link: https://lore.kernel.org/r/20230509084746.48259-1-mirsad.todorovac@alu.unizg… Signed-off-by: Mirsad Todorovac <mirsad.todorovac(a)alu.unizg.hr> [ NOTE: This patch is tested against 5.4 stable. Some of the fixes from the upstream ] [ patch to stable 5.10+ do not apply. ] --- v2: minor clarification and consistency improvements. no change to code. lib/test_firmware.c | 37 ++++++++++++++++++++++++++++--------- 1 file changed, 28 insertions(+), 9 deletions(-) diff --git a/lib/test_firmware.c b/lib/test_firmware.c index 38553944e967..92d7195d5b5b 100644 --- a/lib/test_firmware.c +++ b/lib/test_firmware.c @@ -301,16 +301,26 @@ static ssize_t config_test_show_str(char *dst, return len; } -static int test_dev_config_update_bool(const char *buf, size_t size, - bool *cfg) +static inline int __test_dev_config_update_bool(const char *buf, size_t size, + bool *cfg) { int ret; - mutex_lock(&test_fw_mutex); if (strtobool(buf, cfg) < 0) ret = -EINVAL; else ret = size; + + return ret; +} + +static int test_dev_config_update_bool(const char *buf, size_t size, + bool *cfg) +{ + int ret; + + mutex_lock(&test_fw_mutex); + ret = __test_dev_config_update_bool(buf, size, cfg); mutex_unlock(&test_fw_mutex); return ret; @@ -340,7 +350,7 @@ static ssize_t test_dev_config_show_int(char *buf, int cfg) return snprintf(buf, PAGE_SIZE, "%d\n", val); } -static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) +static inline int __test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) { int ret; long new; @@ -352,14 +362,23 @@ static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) if (new > U8_MAX) return -EINVAL; - mutex_lock(&test_fw_mutex); *(u8 *)cfg = new; - mutex_unlock(&test_fw_mutex); /* Always return full write size even if we didn't consume all */ return size; } +static int test_dev_config_update_u8(const char *buf, size_t size, u8 *cfg) +{ + int ret; + + mutex_lock(&test_fw_mutex); + ret = __test_dev_config_update_u8(buf, size, cfg); + mutex_unlock(&test_fw_mutex); + + return ret; +} + static ssize_t test_dev_config_show_u8(char *buf, u8 cfg) { u8 val; @@ -392,10 +411,10 @@ static ssize_t config_num_requests_store(struct device *dev, mutex_unlock(&test_fw_mutex); goto out; } - mutex_unlock(&test_fw_mutex); - rc = test_dev_config_update_u8(buf, count, - &test_fw_config->num_requests); + rc = __test_dev_config_update_u8(buf, count, + &test_fw_config->num_requests); + mutex_unlock(&test_fw_mutex); out: return rc; -- 2.34.1

1 year, 4 months

1
0
0 0

2024

2023

2022

2021

2020

2019

2018

2017

Linux-kselftest-mirror August 2023