* [PATCH V2 0/1] ltp: upgrade to upstream latest commit on 20170804 @ 2017-08-04 10:00 Dengke Du 2017-08-04 10:00 ` [PATCH V2 1/1] " Dengke Du 0 siblings, 1 reply; 4+ messages in thread From: Dengke Du @ 2017-08-04 10:00 UTC (permalink / raw) To: openembedded-core The following changes since commit 9ed748a542b520c1cb763d981969233c0f5efd4e: bitbake: daemonize: Always print any remaning UI events at exit (2017-08-03 11:14:13 +0100) are available in the git repository at: https://github.com/DengkeDu/openembedded-core.git dengke/ltp-upgrade-to-20170804 https://github.com//tree/dengke/ltp-upgrade-to-20170804 Dengke Du (1): ltp: upgrade to upstream latest commit on 20170804 .../ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch | 34 ------ ...TH_MAX-undeclared-when-building-with-musl.patch | 31 ----- ...faccessat-and-fchmodat-Fix-build-warnings.patch | 68 ----------- ...d_key02-update-to-test-fix-for-nonempty-N.patch | 136 --------------------- meta/recipes-extended/ltp/ltp_20170516.bb | 14 +-- 5 files changed, 1 insertion(+), 282 deletions(-) delete mode 100644 meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch -- 2.7.4 ^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH V2 1/1] ltp: upgrade to upstream latest commit on 20170804 2017-08-04 10:00 [PATCH V2 0/1] ltp: upgrade to upstream latest commit on 20170804 Dengke Du @ 2017-08-04 10:00 ` Dengke Du 2017-08-09 20:47 ` Burton, Ross 0 siblings, 1 reply; 4+ messages in thread From: Dengke Du @ 2017-08-04 10:00 UTC (permalink / raw) To: openembedded-core drop 4 patches because the upstream already contains them: 0001-dirtyc0w-Include-stdint.h.patch 0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch 0037-faccessat-and-fchmodat-Fix-build-warnings.patch 0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch drop the do_compile_prepend function, because the upstream already fix the parallel make race, we can check it here: https://github.com/linux-test-project/ltp/commit/3f385652efe811fe7491474f8513baf44cf0a12d Signed-off-by: Dengke Du <dengke.du@windriver.com> --- .../ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch | 34 ------ ...TH_MAX-undeclared-when-building-with-musl.patch | 31 ----- ...faccessat-and-fchmodat-Fix-build-warnings.patch | 68 ----------- ...d_key02-update-to-test-fix-for-nonempty-N.patch | 136 --------------------- meta/recipes-extended/ltp/ltp_20170516.bb | 14 +-- 5 files changed, 1 insertion(+), 282 deletions(-) delete mode 100644 meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch delete mode 100644 meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch diff --git a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch b/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch deleted file mode 100644 index 5c73be4..0000000 --- a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch +++ /dev/null @@ -1,34 +0,0 @@ -From 1b30d8dac1a37dc0abbb3b545db2824d6489e23f Mon Sep 17 00:00:00 2001 -From: Khem Raj <raj.khem@gmail.com> -Date: Fri, 9 Jun 2017 09:32:13 -0700 -Subject: [PATCH] dirtyc0w: Include stdint.h - -uintptr_t is defined in stdint.h - -Fixes -| dirtyc0w_child.c:76:15: error: 'uintptr_t' undeclared (first use in this function); did you mean 'intptr_t'? -| lseek(mfd, (uintptr_t) map, SEEK_SET); -| ^~~~~~~~~ - -Signed-off-by: Khem Raj <raj.khem@gmail.com> ---- -Upstream-Status: Submitted - - testcases/kernel/security/dirtyc0w/dirtyc0w_child.c | 1 + - 1 file changed, 1 insertion(+) - -diff --git a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c -index 5328a9bc4..49abdd6ba 100644 ---- a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c -+++ b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c -@@ -23,6 +23,7 @@ - #include <sys/stat.h> - #include <string.h> - #include <stdlib.h> -+#include <stdint.h> - #include <pwd.h> - - #include "tst_safe_pthread.h" --- -2.13.1 - diff --git a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch b/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch deleted file mode 100644 index 020ddfe..0000000 --- a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch +++ /dev/null @@ -1,31 +0,0 @@ -From b906ec2fe4711a727e198cd9259287c042eef8e2 Mon Sep 17 00:00:00 2001 -From: Dengke Du <dengke.du@windriver.com> -Date: Thu, 9 Feb 2017 16:41:12 +0800 -Subject: [PATCH] fix PATH_MAX undeclared when building with musl - -fix PATH_MAX undeclared when building with musl. - -Upstream-Status: Submitted [https://github.com/linux-test-project/ltp/pull/176] - -Signed-off-by: Dengke Du <dengke.du@windriver.com> ---- - include/tst_test.h | 3 +++ - 1 file changed, 3 insertions(+) - -diff --git a/include/tst_test.h b/include/tst_test.h -index 335c82684..4dfa1511b 100644 ---- a/include/tst_test.h -+++ b/include/tst_test.h -@@ -23,6 +23,9 @@ - #endif /* __TEST_H__ */ - - #include <unistd.h> -+#ifndef __GLIBC__ -+#include <limits.h> -+#endif - - #include "tst_common.h" - #include "tst_res_flags.h" --- -2.11.0 - diff --git a/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch b/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch deleted file mode 100644 index 99497d6..0000000 --- a/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch +++ /dev/null @@ -1,68 +0,0 @@ -From 4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647 Mon Sep 17 00:00:00 2001 -From: Naresh Kamboju <naresh.kamboju@linaro.org> -Date: Mon, 19 Jun 2017 19:01:21 +0530 -Subject: [PATCH 1/7] faccessat and fchmodat: Fix build warnings -MIME-Version: 1.0 -Content-Type: text/plain; charset=UTF-8 -Content-Transfer-Encoding: 8bit - -Fixing below build warnings. - -ltp-full-20170516/testcases/kernel/syscalls/faccessat$ make -In file included from ../../../../include/old/test.h:51:0, - from faccessat01.c:44: -faccessat01.c: In function ‘setup’: -../../../../include/old/old_safe_file_ops.h:55:27: -warning: format not a string literal and no format arguments [-Wformat-security] - (path), (fmt), ## __VA_ARGS__) - ^ -faccessat01.c:132:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ - SAFE_FILE_PRINTF(cleanup, testfile, testfile); - ^ -../../../../include/old/old_safe_file_ops.h:55:27: -warning: format not a string literal and no format arguments [-Wformat-security] - (path), (fmt), ## __VA_ARGS__) - ^ -faccessat01.c:133:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ - SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); - -Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org> -Upstream-Status: Accepted [https://github.com/linux-test-project/ltp/commit/4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647] ---- - testcases/kernel/syscalls/faccessat/faccessat01.c | 4 ++-- - testcases/kernel/syscalls/fchmodat/fchmodat01.c | 4 ++-- - 2 files changed, 4 insertions(+), 4 deletions(-) - -diff --git a/testcases/kernel/syscalls/faccessat/faccessat01.c b/testcases/kernel/syscalls/faccessat/faccessat01.c -index 622dfd3b0..def7fa755 100644 ---- a/testcases/kernel/syscalls/faccessat/faccessat01.c -+++ b/testcases/kernel/syscalls/faccessat/faccessat01.c -@@ -129,8 +129,8 @@ void setup(void) - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); - fds[1] = fds[4] = fds[0]; - -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); - - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, 0600); - -diff --git a/testcases/kernel/syscalls/fchmodat/fchmodat01.c b/testcases/kernel/syscalls/fchmodat/fchmodat01.c -index 6bf66d8b8..7856ed3fa 100644 ---- a/testcases/kernel/syscalls/fchmodat/fchmodat01.c -+++ b/testcases/kernel/syscalls/fchmodat/fchmodat01.c -@@ -127,8 +127,8 @@ void setup(void) - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); - fds[1] = fds[4] = fds[0]; - -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); - - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, 0600); - fds[3] = 100; --- -2.13.0 - diff --git a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch b/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch deleted file mode 100644 index 3759e1f..0000000 --- a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch +++ /dev/null @@ -1,136 +0,0 @@ -From 25045624e941ee76a13febd36187e23c6c435507 Mon Sep 17 00:00:00 2001 -From: Eric Biggers <ebiggers@google.com> -Date: Mon, 12 Jun 2017 11:55:21 -0700 -Subject: [PATCH 01/39] syscalls/add_key02: update to test fix for nonempty - NULL payload - -add_key02 was supposed to be a "Basic test for the add_key() syscall", -but it actually happened to test the obscure case of passing a NULL -payload with nonzero length. This case was mishandled by the kernel, -which either returned EINVAL or crashed with a NULL pointer dereference, -depending on the key type. (The former applied to the test, as it used -the "user" key type.) The expected behavior in this case is that the -syscall fail with EFAULT. - -Update the test to expect the fixed behavior from v4.12-rc5, and make -the test more thorough by testing additional key types, including ones -that caused a NULL pointer dereference in unfixed kernels. - -Signed-off-by: Eric Biggers <ebiggers@google.com> -Upstream-Status: Accepted [https://github.com/linux-test-project/ltp/commit/25045624e941ee76a13febd36187e23c6c435507] ---- - testcases/kernel/syscalls/add_key/add_key02.c | 69 ++++++++++++++++++++------- - 1 file changed, 51 insertions(+), 18 deletions(-) - -diff --git a/testcases/kernel/syscalls/add_key/add_key02.c b/testcases/kernel/syscalls/add_key/add_key02.c -index 866800d6f..2e3308dcd 100644 ---- a/testcases/kernel/syscalls/add_key/add_key02.c -+++ b/testcases/kernel/syscalls/add_key/add_key02.c -@@ -1,5 +1,6 @@ - /****************************************************************************** - * Copyright (c) Crackerjack Project., 2007 * -+ * Copyright (c) 2017 Google, Inc. * - * * - * This program is free software; you can redistribute it and/or modify * - * it under the terms of the GNU General Public License as published by * -@@ -18,10 +19,17 @@ - ******************************************************************************/ - - /* -- * Basic test for the add_key() syscall. -+ * Test that the add_key() syscall correctly handles a NULL payload with nonzero -+ * length. Specifically, it should fail with EFAULT rather than oopsing the -+ * kernel with a NULL pointer dereference or failing with EINVAL, as it did -+ * before (depending on the key type). This is a regression test for commit -+ * 5649645d725c ("KEYS: fix dereferencing NULL payload with nonzero length"). - * -- * History: Porting from Crackerjack to LTP is done by -- * Manas Kumar Nayak maknayak@in.ibm.com> -+ * Note that none of the key types that exhibited the NULL pointer dereference -+ * are guaranteed to be built into the kernel, so we just test as many as we -+ * can, in the hope of catching one. We also test with the "user" key type for -+ * good measure, although it was one of the types that failed with EINVAL rather -+ * than dereferencing NULL. - */ - - #include "config.h" -@@ -33,36 +41,61 @@ - - #ifdef HAVE_LINUX_KEYCTL_H - struct tcase { -- char *type; -- char *desc; -- void *payload; -- int plen; -- int exp_errno; -+ const char *type; -+ size_t plen; - } tcases[] = { -- {"user", "firstkey", NULL, 1, EINVAL} -+ /* -+ * The payload length we test for each key type needs to pass initial -+ * validation but is otherwise arbitrary. Note: the "rxrpc_s" key type -+ * requires a payload of exactly 8 bytes. -+ */ -+ { "asymmetric", 64 }, -+ { "cifs.idmap", 64 }, -+ { "cifs.spnego", 64 }, -+ { "pkcs7_test", 64 }, -+ { "rxrpc", 64 }, -+ { "rxrpc_s", 8 }, -+ { "user", 64 }, - }; - #endif /* HAVE_LINUX_KEYCTL_H */ - - static void verify_add_key(unsigned int i) - { - #ifdef HAVE_LINUX_KEYCTL_H -- TEST(tst_syscall(__NR_add_key, tcases[i].type, tcases[i].desc, -- tcases[i].payload, tcases[i].plen, -- KEY_SPEC_USER_KEYRING)); -+ TEST(tst_syscall(__NR_add_key, tcases[i].type, "abc:def", -+ NULL, tcases[i].plen, KEY_SPEC_PROCESS_KEYRING)); - - if (TEST_RETURN != -1) { -- tst_res(TFAIL, "add_key() passed unexpectedly"); -+ tst_res(TFAIL, -+ "add_key() with key type '%s' unexpectedly succeeded", -+ tcases[i].type); - return; - } - -- if (TEST_ERRNO == tcases[i].exp_errno) { -- tst_res(TPASS | TTERRNO, "add_key() failed expectedly"); -+ if (TEST_ERRNO == EFAULT) { -+ tst_res(TPASS, "received expected EFAULT with key type '%s'", -+ tcases[i].type); - return; - } - -- tst_res(TFAIL | TTERRNO, -- "add_key() failed unexpectedly, expected %s", -- tst_strerrno(tcases[i].exp_errno)); -+ if (TEST_ERRNO == ENODEV) { -+ tst_res(TCONF, "kernel doesn't support key type '%s'", -+ tcases[i].type); -+ return; -+ } -+ -+ /* -+ * It's possible for the "asymmetric" key type to be supported, but with -+ * no asymmetric key parsers registered. In that case, attempting to -+ * add a key of type asymmetric will fail with EBADMSG. -+ */ -+ if (TEST_ERRNO == EBADMSG && !strcmp(tcases[i].type, "asymmetric")) { -+ tst_res(TCONF, "no asymmetric key parsers are registered"); -+ return; -+ } -+ -+ tst_res(TFAIL | TTERRNO, "unexpected error with key type '%s'", -+ tcases[i].type); - #else - tst_brk(TCONF, "linux/keyctl.h was missing upon compilation."); - #endif /* HAVE_LINUX_KEYCTL_H */ --- -2.13.0 - diff --git a/meta/recipes-extended/ltp/ltp_20170516.bb b/meta/recipes-extended/ltp/ltp_20170516.bb index 86b8012..46c9741 100644 --- a/meta/recipes-extended/ltp/ltp_20170516.bb +++ b/meta/recipes-extended/ltp/ltp_20170516.bb @@ -21,7 +21,7 @@ DEPENDS_append_libc-musl = " fts " EXTRA_OEMAKE_append_libc-musl = " LIBC=musl " CFLAGS_append_powerpc64 = " -D__SANE_USERSPACE_TYPES__" CFLAGS_append_mipsarchn64 = " -D__SANE_USERSPACE_TYPES__" -SRCREV = "6c6c6ca40afb3611e52486f670085762ff451e91" +SRCREV = "18916a2e6d8c997b7b29dcfa9550d5a15b22ed22" SRC_URI = "git://github.com/linux-test-project/ltp.git \ file://0001-add-_GNU_SOURCE-to-pec_listener.c.patch \ @@ -29,7 +29,6 @@ SRC_URI = "git://github.com/linux-test-project/ltp.git \ file://0003-Add-knob-to-control-tirpc-support.patch \ file://0004-build-Add-option-to-select-libc-implementation.patch \ file://0005-kernel-controllers-Link-with-libfts-explicitly-on-mu.patch \ - file://0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch \ file://0007-fix-__WORDSIZE-undeclared-when-building-with-musl.patch \ file://0008-Check-if-__GLIBC_PREREQ-is-defined-before-using-it.patch \ file://0009-fix-redefinition-of-struct-msgbuf-error-building-wit.patch \ @@ -49,9 +48,6 @@ SRC_URI = "git://github.com/linux-test-project/ltp.git \ file://0034-periodic_output.patch \ file://0035-fix-test_proc_kill-hang.patch \ file://0036-testcases-network-nfsv4-acl-acl1.c-Security-fix-on-s.patch \ - file://0001-dirtyc0w-Include-stdint.h.patch \ - file://0037-faccessat-and-fchmodat-Fix-build-warnings.patch \ - file://0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch \ " S = "${WORKDIR}/git" @@ -69,14 +65,6 @@ EXTRA_OECONF = " --with-power-management-testsuite --with-realtime-testsuite " # ltp network/rpc test cases ftbfs when libtirpc is found EXTRA_OECONF += " --without-tirpc " -# The makefiles make excessive use of make -C and several include testcases.mk -# which triggers a build of the syscall header. To reproduce, build ltp, -# then delete the header, then "make -j XX" and watch regen.sh run multiple -# times. Its easier to generate this once here instead. -do_compile_prepend () { - ( make -C ${B}/testcases/kernel include/linux_syscall_numbers.h ) -} - do_install(){ install -d ${D}/opt/ltp/ oe_runmake DESTDIR=${D} SKIP_IDCHECK=1 install -- 2.7.4 ^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [PATCH V2 1/1] ltp: upgrade to upstream latest commit on 20170804 2017-08-04 10:00 ` [PATCH V2 1/1] " Dengke Du @ 2017-08-09 20:47 ` Burton, Ross 2017-08-11 1:15 ` Dengke Du 0 siblings, 1 reply; 4+ messages in thread From: Burton, Ross @ 2017-08-09 20:47 UTC (permalink / raw) To: Dengke Du; +Cc: OE-core [-- Attachment #1: Type: text/plain, Size: 17569 bytes --] Breaks with musl: http://errors.yoctoproject.org/Errors/Details/150199/ Ross On 4 August 2017 at 11:00, Dengke Du <dengke.du@windriver.com> wrote: > drop 4 patches because the upstream already contains them: > > 0001-dirtyc0w-Include-stdint.h.patch > 0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > 0037-faccessat-and-fchmodat-Fix-build-warnings.patch > 0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > > drop the do_compile_prepend function, because the upstream already fix > the parallel make race, we can check it here: > > https://github.com/linux-test-project/ltp/commit/ > 3f385652efe811fe7491474f8513baf44cf0a12d > > Signed-off-by: Dengke Du <dengke.du@windriver.com> > --- > .../ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch | 34 ------ > ...TH_MAX-undeclared-when-building-with-musl.patch | 31 ----- > ...faccessat-and-fchmodat-Fix-build-warnings.patch | 68 ----------- > ...d_key02-update-to-test-fix-for-nonempty-N.patch | 136 > --------------------- > meta/recipes-extended/ltp/ltp_20170516.bb | 14 +-- > 5 files changed, 1 insertion(+), 282 deletions(-) > delete mode 100644 meta/recipes-extended/ltp/ltp/ > 0001-dirtyc0w-Include-stdint.h.patch > delete mode 100644 meta/recipes-extended/ltp/ltp/ > 0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > delete mode 100644 meta/recipes-extended/ltp/ltp/ > 0037-faccessat-and-fchmodat-Fix-build-warnings.patch > delete mode 100644 meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02- > update-to-test-fix-for-nonempty-N.patch > > diff --git a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > b/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > deleted file mode 100644 > index 5c73be4..0000000 > --- a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > +++ /dev/null > @@ -1,34 +0,0 @@ > -From 1b30d8dac1a37dc0abbb3b545db2824d6489e23f Mon Sep 17 00:00:00 2001 > -From: Khem Raj <raj.khem@gmail.com> > -Date: Fri, 9 Jun 2017 09:32:13 -0700 > -Subject: [PATCH] dirtyc0w: Include stdint.h > - > -uintptr_t is defined in stdint.h > - > -Fixes > -| dirtyc0w_child.c:76:15: error: 'uintptr_t' undeclared (first use in > this function); did you mean 'intptr_t'? > -| lseek(mfd, (uintptr_t) map, SEEK_SET); > -| ^~~~~~~~~ > - > -Signed-off-by: Khem Raj <raj.khem@gmail.com> > ---- > -Upstream-Status: Submitted > - > - testcases/kernel/security/dirtyc0w/dirtyc0w_child.c | 1 + > - 1 file changed, 1 insertion(+) > - > -diff --git a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -index 5328a9bc4..49abdd6ba 100644 > ---- a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -+++ b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -@@ -23,6 +23,7 @@ > - #include <sys/stat.h> > - #include <string.h> > - #include <stdlib.h> > -+#include <stdint.h> > - #include <pwd.h> > - > - #include "tst_safe_pthread.h" > --- > -2.13.1 > - > diff --git a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX- > undeclared-when-building-with-musl.patch b/meta/recipes-extended/ltp/ > ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > deleted file mode 100644 > index 020ddfe..0000000 > --- a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX- > undeclared-when-building-with-musl.patch > +++ /dev/null > @@ -1,31 +0,0 @@ > -From b906ec2fe4711a727e198cd9259287c042eef8e2 Mon Sep 17 00:00:00 2001 > -From: Dengke Du <dengke.du@windriver.com> > -Date: Thu, 9 Feb 2017 16:41:12 +0800 > -Subject: [PATCH] fix PATH_MAX undeclared when building with musl > - > -fix PATH_MAX undeclared when building with musl. > - > -Upstream-Status: Submitted [https://github.com/linux- > test-project/ltp/pull/176] > - > -Signed-off-by: Dengke Du <dengke.du@windriver.com> > ---- > - include/tst_test.h | 3 +++ > - 1 file changed, 3 insertions(+) > - > -diff --git a/include/tst_test.h b/include/tst_test.h > -index 335c82684..4dfa1511b 100644 > ---- a/include/tst_test.h > -+++ b/include/tst_test.h > -@@ -23,6 +23,9 @@ > - #endif /* __TEST_H__ */ > - > - #include <unistd.h> > -+#ifndef __GLIBC__ > -+#include <limits.h> > -+#endif > - > - #include "tst_common.h" > - #include "tst_res_flags.h" > --- > -2.11.0 > - > diff --git a/meta/recipes-extended/ltp/ltp/0037-faccessat-and- > fchmodat-Fix-build-warnings.patch b/meta/recipes-extended/ltp/ > ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch > deleted file mode 100644 > index 99497d6..0000000 > --- a/meta/recipes-extended/ltp/ltp/0037-faccessat-and- > fchmodat-Fix-build-warnings.patch > +++ /dev/null > @@ -1,68 +0,0 @@ > -From 4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647 Mon Sep 17 00:00:00 2001 > -From: Naresh Kamboju <naresh.kamboju@linaro.org> > -Date: Mon, 19 Jun 2017 19:01:21 +0530 > -Subject: [PATCH 1/7] faccessat and fchmodat: Fix build warnings > -MIME-Version: 1.0 > -Content-Type: text/plain; charset=UTF-8 > -Content-Transfer-Encoding: 8bit > - > -Fixing below build warnings. > - > -ltp-full-20170516/testcases/kernel/syscalls/faccessat$ make > -In file included from ../../../../include/old/test.h:51:0, > - from faccessat01.c:44: > -faccessat01.c: In function ‘setup’: > -../../../../include/old/old_safe_file_ops.h:55:27: > -warning: format not a string literal and no format arguments > [-Wformat-security] > - (path), (fmt), ## __VA_ARGS__) > - ^ > -faccessat01.c:132:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ > - SAFE_FILE_PRINTF(cleanup, testfile, testfile); > - ^ > -../../../../include/old/old_safe_file_ops.h:55:27: > -warning: format not a string literal and no format arguments > [-Wformat-security] > - (path), (fmt), ## __VA_ARGS__) > - ^ > -faccessat01.c:133:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ > - SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > - > -Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org> > -Upstream-Status: Accepted [https://github.com/linux- > test-project/ltp/commit/4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647] > ---- > - testcases/kernel/syscalls/faccessat/faccessat01.c | 4 ++-- > - testcases/kernel/syscalls/fchmodat/fchmodat01.c | 4 ++-- > - 2 files changed, 4 insertions(+), 4 deletions(-) > - > -diff --git a/testcases/kernel/syscalls/faccessat/faccessat01.c > b/testcases/kernel/syscalls/faccessat/faccessat01.c > -index 622dfd3b0..def7fa755 100644 > ---- a/testcases/kernel/syscalls/faccessat/faccessat01.c > -+++ b/testcases/kernel/syscalls/faccessat/faccessat01.c > -@@ -129,8 +129,8 @@ void setup(void) > - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); > - fds[1] = fds[4] = fds[0]; > - > -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); > -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); > -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); > - > - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, 0600); > - > -diff --git a/testcases/kernel/syscalls/fchmodat/fchmodat01.c > b/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -index 6bf66d8b8..7856ed3fa 100644 > ---- a/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -+++ b/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -@@ -127,8 +127,8 @@ void setup(void) > - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); > - fds[1] = fds[4] = fds[0]; > - > -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); > -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); > -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); > - > - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, 0600); > - fds[3] = 100; > --- > -2.13.0 > - > diff --git a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02- > update-to-test-fix-for-nonempty-N.patch b/meta/recipes-extended/ltp/ > ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > deleted file mode 100644 > index 3759e1f..0000000 > --- a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02- > update-to-test-fix-for-nonempty-N.patch > +++ /dev/null > @@ -1,136 +0,0 @@ > -From 25045624e941ee76a13febd36187e23c6c435507 Mon Sep 17 00:00:00 2001 > -From: Eric Biggers <ebiggers@google.com> > -Date: Mon, 12 Jun 2017 11:55:21 -0700 > -Subject: [PATCH 01/39] syscalls/add_key02: update to test fix for nonempty > - NULL payload > - > -add_key02 was supposed to be a "Basic test for the add_key() syscall", > -but it actually happened to test the obscure case of passing a NULL > -payload with nonzero length. This case was mishandled by the kernel, > -which either returned EINVAL or crashed with a NULL pointer dereference, > -depending on the key type. (The former applied to the test, as it used > -the "user" key type.) The expected behavior in this case is that the > -syscall fail with EFAULT. > - > -Update the test to expect the fixed behavior from v4.12-rc5, and make > -the test more thorough by testing additional key types, including ones > -that caused a NULL pointer dereference in unfixed kernels. > - > -Signed-off-by: Eric Biggers <ebiggers@google.com> > -Upstream-Status: Accepted [https://github.com/linux- > test-project/ltp/commit/25045624e941ee76a13febd36187e23c6c435507] > ---- > - testcases/kernel/syscalls/add_key/add_key02.c | 69 > ++++++++++++++++++++------- > - 1 file changed, 51 insertions(+), 18 deletions(-) > - > -diff --git a/testcases/kernel/syscalls/add_key/add_key02.c > b/testcases/kernel/syscalls/add_key/add_key02.c > -index 866800d6f..2e3308dcd 100644 > ---- a/testcases/kernel/syscalls/add_key/add_key02.c > -+++ b/testcases/kernel/syscalls/add_key/add_key02.c > -@@ -1,5 +1,6 @@ > - /*********************************************************** > ******************* > - * Copyright (c) Crackerjack Project., 2007 > * > -+ * Copyright (c) 2017 Google, Inc. > * > - * > * > - * This program is free software; you can redistribute it and/or > modify * > - * it under the terms of the GNU General Public License as published by > * > -@@ -18,10 +19,17 @@ > - ************************************************************ > ******************/ > - > - /* > -- * Basic test for the add_key() syscall. > -+ * Test that the add_key() syscall correctly handles a NULL payload with > nonzero > -+ * length. Specifically, it should fail with EFAULT rather than oopsing > the > -+ * kernel with a NULL pointer dereference or failing with EINVAL, as it > did > -+ * before (depending on the key type). This is a regression test for > commit > -+ * 5649645d725c ("KEYS: fix dereferencing NULL payload with nonzero > length"). > - * > -- * History: Porting from Crackerjack to LTP is done by > -- * Manas Kumar Nayak maknayak@in.ibm.com> > -+ * Note that none of the key types that exhibited the NULL pointer > dereference > -+ * are guaranteed to be built into the kernel, so we just test as many > as we > -+ * can, in the hope of catching one. We also test with the "user" key > type for > -+ * good measure, although it was one of the types that failed with > EINVAL rather > -+ * than dereferencing NULL. > - */ > - > - #include "config.h" > -@@ -33,36 +41,61 @@ > - > - #ifdef HAVE_LINUX_KEYCTL_H > - struct tcase { > -- char *type; > -- char *desc; > -- void *payload; > -- int plen; > -- int exp_errno; > -+ const char *type; > -+ size_t plen; > - } tcases[] = { > -- {"user", "firstkey", NULL, 1, EINVAL} > -+ /* > -+ * The payload length we test for each key type needs to pass > initial > -+ * validation but is otherwise arbitrary. Note: the "rxrpc_s" key > type > -+ * requires a payload of exactly 8 bytes. > -+ */ > -+ { "asymmetric", 64 }, > -+ { "cifs.idmap", 64 }, > -+ { "cifs.spnego", 64 }, > -+ { "pkcs7_test", 64 }, > -+ { "rxrpc", 64 }, > -+ { "rxrpc_s", 8 }, > -+ { "user", 64 }, > - }; > - #endif /* HAVE_LINUX_KEYCTL_H */ > - > - static void verify_add_key(unsigned int i) > - { > - #ifdef HAVE_LINUX_KEYCTL_H > -- TEST(tst_syscall(__NR_add_key, tcases[i].type, tcases[i].desc, > -- tcases[i].payload, tcases[i].plen, > -- KEY_SPEC_USER_KEYRING)); > -+ TEST(tst_syscall(__NR_add_key, tcases[i].type, "abc:def", > -+ NULL, tcases[i].plen, KEY_SPEC_PROCESS_KEYRING)); > - > - if (TEST_RETURN != -1) { > -- tst_res(TFAIL, "add_key() passed unexpectedly"); > -+ tst_res(TFAIL, > -+ "add_key() with key type '%s' unexpectedly > succeeded", > -+ tcases[i].type); > - return; > - } > - > -- if (TEST_ERRNO == tcases[i].exp_errno) { > -- tst_res(TPASS | TTERRNO, "add_key() failed expectedly"); > -+ if (TEST_ERRNO == EFAULT) { > -+ tst_res(TPASS, "received expected EFAULT with key type > '%s'", > -+ tcases[i].type); > - return; > - } > - > -- tst_res(TFAIL | TTERRNO, > -- "add_key() failed unexpectedly, expected %s", > -- tst_strerrno(tcases[i].exp_errno)); > -+ if (TEST_ERRNO == ENODEV) { > -+ tst_res(TCONF, "kernel doesn't support key type '%s'", > -+ tcases[i].type); > -+ return; > -+ } > -+ > -+ /* > -+ * It's possible for the "asymmetric" key type to be supported, > but with > -+ * no asymmetric key parsers registered. In that case, attempting > to > -+ * add a key of type asymmetric will fail with EBADMSG. > -+ */ > -+ if (TEST_ERRNO == EBADMSG && !strcmp(tcases[i].type, > "asymmetric")) { > -+ tst_res(TCONF, "no asymmetric key parsers are registered"); > -+ return; > -+ } > -+ > -+ tst_res(TFAIL | TTERRNO, "unexpected error with key type '%s'", > -+ tcases[i].type); > - #else > - tst_brk(TCONF, "linux/keyctl.h was missing upon compilation."); > - #endif /* HAVE_LINUX_KEYCTL_H */ > --- > -2.13.0 > - > diff --git a/meta/recipes-extended/ltp/ltp_20170516.bb > b/meta/recipes-extended/ltp/ltp_20170516.bb > index 86b8012..46c9741 100644 > --- a/meta/recipes-extended/ltp/ltp_20170516.bb > +++ b/meta/recipes-extended/ltp/ltp_20170516.bb > @@ -21,7 +21,7 @@ DEPENDS_append_libc-musl = " fts " > EXTRA_OEMAKE_append_libc-musl = " LIBC=musl " > CFLAGS_append_powerpc64 = " -D__SANE_USERSPACE_TYPES__" > CFLAGS_append_mipsarchn64 = " -D__SANE_USERSPACE_TYPES__" > -SRCREV = "6c6c6ca40afb3611e52486f670085762ff451e91" > +SRCREV = "18916a2e6d8c997b7b29dcfa9550d5a15b22ed22" > > SRC_URI = "git://github.com/linux-test-project/ltp.git \ > file://0001-add-_GNU_SOURCE-to-pec_listener.c.patch \ > @@ -29,7 +29,6 @@ SRC_URI = "git://github.com/linux-test-project/ltp.git \ > file://0003-Add-knob-to-control-tirpc-support.patch \ > file://0004-build-Add-option-to-select-libc-implementation.patch > \ > file://0005-kernel-controllers-Link-with-libfts-explicitly-on-mu.patch > \ > - file://0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > \ > file://0007-fix-__WORDSIZE-undeclared-when-building-with-musl.patch > \ > file://0008-Check-if-__GLIBC_PREREQ-is-defined-before-using-it.patch > \ > file://0009-fix-redefinition-of-struct-msgbuf-error-building-wit.patch > \ > @@ -49,9 +48,6 @@ SRC_URI = "git://github.com/linux-test-project/ltp.git \ > file://0034-periodic_output.patch \ > file://0035-fix-test_proc_kill-hang.patch \ > file://0036-testcases-network-nfsv4-acl-acl1.c-Security-fix-on-s.patch > \ > - file://0001-dirtyc0w-Include-stdint.h.patch \ > - file://0037-faccessat-and-fchmodat-Fix-build-warnings.patch \ > - file://0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > \ > " > > S = "${WORKDIR}/git" > @@ -69,14 +65,6 @@ EXTRA_OECONF = " --with-power-management-testsuite > --with-realtime-testsuite " > # ltp network/rpc test cases ftbfs when libtirpc is found > EXTRA_OECONF += " --without-tirpc " > > -# The makefiles make excessive use of make -C and several include > testcases.mk > -# which triggers a build of the syscall header. To reproduce, build ltp, > -# then delete the header, then "make -j XX" and watch regen.sh run > multiple > -# times. Its easier to generate this once here instead. > -do_compile_prepend () { > - ( make -C ${B}/testcases/kernel include/linux_syscall_numbers.h ) > -} > - > do_install(){ > install -d ${D}/opt/ltp/ > oe_runmake DESTDIR=${D} SKIP_IDCHECK=1 install > -- > 2.7.4 > > -- > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > http://lists.openembedded.org/mailman/listinfo/openembedded-core > [-- Attachment #2: Type: text/html, Size: 22535 bytes --] ^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH V2 1/1] ltp: upgrade to upstream latest commit on 20170804 2017-08-09 20:47 ` Burton, Ross @ 2017-08-11 1:15 ` Dengke Du 0 siblings, 0 replies; 4+ messages in thread From: Dengke Du @ 2017-08-11 1:15 UTC (permalink / raw) To: Burton, Ross; +Cc: OE-core [-- Attachment #1: Type: text/plain, Size: 20891 bytes --] OK, I have send the fix when building with musl for the upstream: https://github.com/linux-test-project/ltp/pull/194 still waitting. On 2017年08月10日 04:47, Burton, Ross wrote: > Breaks with musl: > > http://errors.yoctoproject.org/Errors/Details/150199/ > > Ross > > On 4 August 2017 at 11:00, Dengke Du <dengke.du@windriver.com > <mailto:dengke.du@windriver.com>> wrote: > > drop 4 patches because the upstream already contains them: > > 0001-dirtyc0w-Include-stdint.h.patch > 0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > 0037-faccessat-and-fchmodat-Fix-build-warnings.patch > 0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > > drop the do_compile_prepend function, because the upstream already fix > the parallel make race, we can check it here: > > https://github.com/linux-test-project/ltp/commit/3f385652efe811fe7491474f8513baf44cf0a12d > <https://github.com/linux-test-project/ltp/commit/3f385652efe811fe7491474f8513baf44cf0a12d> > > Signed-off-by: Dengke Du <dengke.du@windriver.com > <mailto:dengke.du@windriver.com>> > --- > .../ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch | 34 ------ > ...TH_MAX-undeclared-when-building-with-musl.patch | 31 ----- > ...faccessat-and-fchmodat-Fix-build-warnings.patch | 68 ----------- > ...d_key02-update-to-test-fix-for-nonempty-N.patch | 136 > --------------------- > meta/recipes-extended/ltp/ltp_20170516.bb > <http://ltp_20170516.bb> | 14 +-- > 5 files changed, 1 insertion(+), 282 deletions(-) > delete mode 100644 > meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > delete mode 100644 > meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > delete mode 100644 > meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch > delete mode 100644 > meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > > diff --git > a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > b/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > deleted file mode 100644 > index 5c73be4..0000000 > --- > a/meta/recipes-extended/ltp/ltp/0001-dirtyc0w-Include-stdint.h.patch > +++ /dev/null > @@ -1,34 +0,0 @@ > -From 1b30d8dac1a37dc0abbb3b545db2824d6489e23f Mon Sep 17 00:00:00 > 2001 > -From: Khem Raj <raj.khem@gmail.com <mailto:raj.khem@gmail.com>> > -Date: Fri, 9 Jun 2017 09:32:13 -0700 > -Subject: [PATCH] dirtyc0w: Include stdint.h > - > -uintptr_t is defined in stdint.h > - > -Fixes > -| dirtyc0w_child.c:76:15: error: 'uintptr_t' undeclared (first > use in this function); did you mean 'intptr_t'? > -| lseek(mfd, (uintptr_t) map, SEEK_SET); > -| ^~~~~~~~~ > - > -Signed-off-by: Khem Raj <raj.khem@gmail.com > <mailto:raj.khem@gmail.com>> > ---- > -Upstream-Status: Submitted > - > - testcases/kernel/security/dirtyc0w/dirtyc0w_child.c | 1 + > - 1 file changed, 1 insertion(+) > - > -diff --git a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -index 5328a9bc4..49abdd6ba 100644 > ---- a/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -+++ b/testcases/kernel/security/dirtyc0w/dirtyc0w_child.c > -@@ -23,6 +23,7 @@ > - #include <sys/stat.h> > - #include <string.h> > - #include <stdlib.h> > -+#include <stdint.h> > - #include <pwd.h> > - > - #include "tst_safe_pthread.h" > --- > -2.13.1 > - > diff --git > a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > b/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > deleted file mode 100644 > index 020ddfe..0000000 > --- > a/meta/recipes-extended/ltp/ltp/0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch > +++ /dev/null > @@ -1,31 +0,0 @@ > -From b906ec2fe4711a727e198cd9259287c042eef8e2 Mon Sep 17 00:00:00 > 2001 > -From: Dengke Du <dengke.du@windriver.com > <mailto:dengke.du@windriver.com>> > -Date: Thu, 9 Feb 2017 16:41:12 +0800 > -Subject: [PATCH] fix PATH_MAX undeclared when building with musl > - > -fix PATH_MAX undeclared when building with musl. > - > -Upstream-Status: Submitted > [https://github.com/linux-test-project/ltp/pull/176 > <https://github.com/linux-test-project/ltp/pull/176>] > - > -Signed-off-by: Dengke Du <dengke.du@windriver.com > <mailto:dengke.du@windriver.com>> > ---- > - include/tst_test.h | 3 +++ > - 1 file changed, 3 insertions(+) > - > -diff --git a/include/tst_test.h b/include/tst_test.h > -index 335c82684..4dfa1511b 100644 > ---- a/include/tst_test.h > -+++ b/include/tst_test.h > -@@ -23,6 +23,9 @@ > - #endif /* __TEST_H__ */ > - > - #include <unistd.h> > -+#ifndef __GLIBC__ > -+#include <limits.h> > -+#endif > - > - #include "tst_common.h" > - #include "tst_res_flags.h" > --- > -2.11.0 > - > diff --git > a/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch > b/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch > deleted file mode 100644 > index 99497d6..0000000 > --- > a/meta/recipes-extended/ltp/ltp/0037-faccessat-and-fchmodat-Fix-build-warnings.patch > +++ /dev/null > @@ -1,68 +0,0 @@ > -From 4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647 Mon Sep 17 00:00:00 > 2001 > -From: Naresh Kamboju <naresh.kamboju@linaro.org > <mailto:naresh.kamboju@linaro.org>> > -Date: Mon, 19 Jun 2017 19:01:21 +0530 > -Subject: [PATCH 1/7] faccessat and fchmodat: Fix build warnings > -MIME-Version: 1.0 > -Content-Type: text/plain; charset=UTF-8 > -Content-Transfer-Encoding: 8bit > - > -Fixing below build warnings. > - > -ltp-full-20170516/testcases/kernel/syscalls/faccessat$ make > -In file included from ../../../../include/old/test.h:51:0, > - from faccessat01.c:44: > -faccessat01.c: In function ‘setup’: > -../../../../include/old/old_safe_file_ops.h:55:27: > -warning: format not a string literal and no format arguments > [-Wformat-security] > - (path), (fmt), ## __VA_ARGS__) > - ^ > -faccessat01.c:132:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ > - SAFE_FILE_PRINTF(cleanup, testfile, testfile); > - ^ > -../../../../include/old/old_safe_file_ops.h:55:27: > -warning: format not a string literal and no format arguments > [-Wformat-security] > - (path), (fmt), ## __VA_ARGS__) > - ^ > -faccessat01.c:133:2: note: in expansion of macro ‘SAFE_FILE_PRINTF’ > - SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > - > -Signed-off-by: Naresh Kamboju <naresh.kamboju@linaro.org > <mailto:naresh.kamboju@linaro.org>> > -Upstream-Status: Accepted > [https://github.com/linux-test-project/ltp/commit/4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647 > <https://github.com/linux-test-project/ltp/commit/4141bdb441f7576a2e73abeb2a0fb1c7b8bd8647>] > ---- > - testcases/kernel/syscalls/faccessat/faccessat01.c | 4 ++-- > - testcases/kernel/syscalls/fchmodat/fchmodat01.c | 4 ++-- > - 2 files changed, 4 insertions(+), 4 deletions(-) > - > -diff --git a/testcases/kernel/syscalls/faccessat/faccessat01.c > b/testcases/kernel/syscalls/faccessat/faccessat01.c > -index 622dfd3b0..def7fa755 100644 > ---- a/testcases/kernel/syscalls/faccessat/faccessat01.c > -+++ b/testcases/kernel/syscalls/faccessat/faccessat01.c > -@@ -129,8 +129,8 @@ void setup(void) > - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); > - fds[1] = fds[4] = fds[0]; > - > -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); > -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); > -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); > - > - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, > 0600); > - > -diff --git a/testcases/kernel/syscalls/fchmodat/fchmodat01.c > b/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -index 6bf66d8b8..7856ed3fa 100644 > ---- a/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -+++ b/testcases/kernel/syscalls/fchmodat/fchmodat01.c > -@@ -127,8 +127,8 @@ void setup(void) > - fds[0] = SAFE_OPEN(cleanup, pathname, O_DIRECTORY); > - fds[1] = fds[4] = fds[0]; > - > -- SAFE_FILE_PRINTF(cleanup, testfile, testfile); > -- SAFE_FILE_PRINTF(cleanup, testfile2, testfile2); > -+ SAFE_FILE_PRINTF(cleanup, testfile, "%s", testfile); > -+ SAFE_FILE_PRINTF(cleanup, testfile2, "%s", testfile2); > - > - fds[2] = SAFE_OPEN(cleanup, testfile3, O_CREAT | O_RDWR, > 0600); > - fds[3] = 100; > --- > -2.13.0 > - > diff --git > a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > b/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > deleted file mode 100644 > index 3759e1f..0000000 > --- > a/meta/recipes-extended/ltp/ltp/0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > +++ /dev/null > @@ -1,136 +0,0 @@ > -From 25045624e941ee76a13febd36187e23c6c435507 Mon Sep 17 00:00:00 > 2001 > -From: Eric Biggers <ebiggers@google.com <mailto:ebiggers@google.com>> > -Date: Mon, 12 Jun 2017 11:55:21 -0700 > -Subject: [PATCH 01/39] syscalls/add_key02: update to test fix for > nonempty > - NULL payload > - > -add_key02 was supposed to be a "Basic test for the add_key() > syscall", > -but it actually happened to test the obscure case of passing a NULL > -payload with nonzero length. This case was mishandled by the kernel, > -which either returned EINVAL or crashed with a NULL pointer > dereference, > -depending on the key type. (The former applied to the test, as > it used > -the "user" key type.) The expected behavior in this case is that the > -syscall fail with EFAULT. > - > -Update the test to expect the fixed behavior from v4.12-rc5, and make > -the test more thorough by testing additional key types, including > ones > -that caused a NULL pointer dereference in unfixed kernels. > - > -Signed-off-by: Eric Biggers <ebiggers@google.com > <mailto:ebiggers@google.com>> > -Upstream-Status: Accepted > [https://github.com/linux-test-project/ltp/commit/25045624e941ee76a13febd36187e23c6c435507 > <https://github.com/linux-test-project/ltp/commit/25045624e941ee76a13febd36187e23c6c435507>] > ---- > - testcases/kernel/syscalls/add_key/add_key02.c | 69 > ++++++++++++++++++++------- > - 1 file changed, 51 insertions(+), 18 deletions(-) > - > -diff --git a/testcases/kernel/syscalls/add_key/add_key02.c > b/testcases/kernel/syscalls/add_key/add_key02.c > -index 866800d6f..2e3308dcd 100644 > ---- a/testcases/kernel/syscalls/add_key/add_key02.c > -+++ b/testcases/kernel/syscalls/add_key/add_key02.c > -@@ -1,5 +1,6 @@ > - > /****************************************************************************** > - * Copyright (c) Crackerjack Project., 2007 * > -+ * Copyright (c) 2017 Google, Inc. * > - * * > - * This program is free software; you can redistribute it > and/or modify * > - * it under the terms of the GNU General Public License as > published by * > -@@ -18,10 +19,17 @@ > - > ******************************************************************************/ > - > - /* > -- * Basic test for the add_key() syscall. > -+ * Test that the add_key() syscall correctly handles a NULL > payload with nonzero > -+ * length. Specifically, it should fail with EFAULT rather than > oopsing the > -+ * kernel with a NULL pointer dereference or failing with > EINVAL, as it did > -+ * before (depending on the key type). This is a regression > test for commit > -+ * 5649645d725c ("KEYS: fix dereferencing NULL payload with > nonzero length"). > - * > -- * History: Porting from Crackerjack to LTP is done by > -- * Manas Kumar Nayak maknayak@in.ibm.com > <mailto:maknayak@in.ibm.com>> > -+ * Note that none of the key types that exhibited the NULL > pointer dereference > -+ * are guaranteed to be built into the kernel, so we just test > as many as we > -+ * can, in the hope of catching one. We also test with the > "user" key type for > -+ * good measure, although it was one of the types that failed > with EINVAL rather > -+ * than dereferencing NULL. > - */ > - > - #include "config.h" > -@@ -33,36 +41,61 @@ > - > - #ifdef HAVE_LINUX_KEYCTL_H > - struct tcase { > -- char *type; > -- char *desc; > -- void *payload; > -- int plen; > -- int exp_errno; > -+ const char *type; > -+ size_t plen; > - } tcases[] = { > -- {"user", "firstkey", NULL, 1, EINVAL} > -+ /* > -+ * The payload length we test for each key type needs to > pass initial > -+ * validation but is otherwise arbitrary. Note: the > "rxrpc_s" key type > -+ * requires a payload of exactly 8 bytes. > -+ */ > -+ { "asymmetric", 64 }, > -+ { "cifs.idmap", 64 }, > -+ { "cifs.spnego", 64 }, > -+ { "pkcs7_test", 64 }, > -+ { "rxrpc", 64 }, > -+ { "rxrpc_s", 8 }, > -+ { "user", 64 }, > - }; > - #endif /* HAVE_LINUX_KEYCTL_H */ > - > - static void verify_add_key(unsigned int i) > - { > - #ifdef HAVE_LINUX_KEYCTL_H > -- TEST(tst_syscall(__NR_add_key, tcases[i].type, tcases[i].desc, > -- tcases[i].payload, tcases[i].plen, > -- KEY_SPEC_USER_KEYRING)); > -+ TEST(tst_syscall(__NR_add_key, tcases[i].type, "abc:def", > -+ NULL, tcases[i].plen, > KEY_SPEC_PROCESS_KEYRING)); > - > - if (TEST_RETURN != -1) { > -- tst_res(TFAIL, "add_key() passed unexpectedly"); > -+ tst_res(TFAIL, > -+ "add_key() with key type '%s' unexpectedly > succeeded", > -+ tcases[i].type); > - return; > - } > - > -- if (TEST_ERRNO == tcases[i].exp_errno) { > -- tst_res(TPASS | TTERRNO, "add_key() failed > expectedly"); > -+ if (TEST_ERRNO == EFAULT) { > -+ tst_res(TPASS, "received expected EFAULT with key > type '%s'", > -+ tcases[i].type); > - return; > - } > - > -- tst_res(TFAIL | TTERRNO, > -- "add_key() failed unexpectedly, expected %s", > -- tst_strerrno(tcases[i].exp_errno)); > -+ if (TEST_ERRNO == ENODEV) { > -+ tst_res(TCONF, "kernel doesn't support key type '%s'", > -+ tcases[i].type); > -+ return; > -+ } > -+ > -+ /* > -+ * It's possible for the "asymmetric" key type to be > supported, but with > -+ * no asymmetric key parsers registered. In that case, > attempting to > -+ * add a key of type asymmetric will fail with EBADMSG. > -+ */ > -+ if (TEST_ERRNO == EBADMSG && !strcmp(tcases[i].type, > "asymmetric")) { > -+ tst_res(TCONF, "no asymmetric key parsers are > registered"); > -+ return; > -+ } > -+ > -+ tst_res(TFAIL | TTERRNO, "unexpected error with key type > '%s'", > -+ tcases[i].type); > - #else > - tst_brk(TCONF, "linux/keyctl.h was missing upon > compilation."); > - #endif /* HAVE_LINUX_KEYCTL_H */ > --- > -2.13.0 > - > diff --git a/meta/recipes-extended/ltp/ltp_20170516.bb > <http://ltp_20170516.bb> > b/meta/recipes-extended/ltp/ltp_20170516.bb <http://ltp_20170516.bb> > index 86b8012..46c9741 100644 > --- a/meta/recipes-extended/ltp/ltp_20170516.bb > <http://ltp_20170516.bb> > +++ b/meta/recipes-extended/ltp/ltp_20170516.bb > <http://ltp_20170516.bb> > @@ -21,7 +21,7 @@ DEPENDS_append_libc-musl = " fts " > EXTRA_OEMAKE_append_libc-musl = " LIBC=musl " > CFLAGS_append_powerpc64 = " -D__SANE_USERSPACE_TYPES__" > CFLAGS_append_mipsarchn64 = " -D__SANE_USERSPACE_TYPES__" > -SRCREV = "6c6c6ca40afb3611e52486f670085762ff451e91" > +SRCREV = "18916a2e6d8c997b7b29dcfa9550d5a15b22ed22" > > SRC_URI = "git://github.com/linux-test-project/ltp.git > <http://github.com/linux-test-project/ltp.git> \ > file://0001-add-_GNU_SOURCE-to-pec_listener.c.patch \ > @@ -29,7 +29,6 @@ SRC_URI = > "git://github.com/linux-test-project/ltp.git > <http://github.com/linux-test-project/ltp.git> \ > file://0003-Add-knob-to-control-tirpc-support.patch \ > > file://0004-build-Add-option-to-select-libc-implementation.patch \ > > file://0005-kernel-controllers-Link-with-libfts-explicitly-on-mu.patch > \ > - > file://0006-fix-PATH_MAX-undeclared-when-building-with-musl.patch \ > > file://0007-fix-__WORDSIZE-undeclared-when-building-with-musl.patch \ > > file://0008-Check-if-__GLIBC_PREREQ-is-defined-before-using-it.patch \ > > file://0009-fix-redefinition-of-struct-msgbuf-error-building-wit.patch > \ > @@ -49,9 +48,6 @@ SRC_URI = > "git://github.com/linux-test-project/ltp.git > <http://github.com/linux-test-project/ltp.git> \ > file://0034-periodic_output.patch \ > file://0035-fix-test_proc_kill-hang.patch \ > > file://0036-testcases-network-nfsv4-acl-acl1.c-Security-fix-on-s.patch > \ > - file://0001-dirtyc0w-Include-stdint.h.patch \ > - > file://0037-faccessat-and-fchmodat-Fix-build-warnings.patch \ > - > file://0038-syscalls-add_key02-update-to-test-fix-for-nonempty-N.patch > \ > " > > S = "${WORKDIR}/git" > @@ -69,14 +65,6 @@ EXTRA_OECONF = " > --with-power-management-testsuite --with-realtime-testsuite " > # ltp network/rpc test cases ftbfs when libtirpc is found > EXTRA_OECONF += " --without-tirpc " > > -# The makefiles make excessive use of make -C and several include > testcases.mk <http://testcases.mk> > -# which triggers a build of the syscall header. To reproduce, > build ltp, > -# then delete the header, then "make -j XX" and watch regen.sh > run multiple > -# times. Its easier to generate this once here instead. > -do_compile_prepend () { > - ( make -C ${B}/testcases/kernel > include/linux_syscall_numbers.h ) > -} > - > do_install(){ > install -d ${D}/opt/ltp/ > oe_runmake DESTDIR=${D} SKIP_IDCHECK=1 install > -- > 2.7.4 > > -- > _______________________________________________ > Openembedded-core mailing list > Openembedded-core@lists.openembedded.org > <mailto:Openembedded-core@lists.openembedded.org> > http://lists.openembedded.org/mailman/listinfo/openembedded-core > <http://lists.openembedded.org/mailman/listinfo/openembedded-core> > > [-- Attachment #2: Type: text/html, Size: 31005 bytes --] ^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2017-08-11 1:16 UTC | newest] Thread overview: 4+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2017-08-04 10:00 [PATCH V2 0/1] ltp: upgrade to upstream latest commit on 20170804 Dengke Du 2017-08-04 10:00 ` [PATCH V2 1/1] " Dengke Du 2017-08-09 20:47 ` Burton, Ross 2017-08-11 1:15 ` Dengke Du
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox