)]}' { "log": [ { "commit": "275aaa68334c45f616c6576f20201e9890c5da18", "tree": "a00b6a37ccdc823567a0a1ae414b7275c33167b6", "parents": [ "326b1ffc136d944d403a2b44dc3abace2a4e10fd" ], "author": { "name": "Arnd Bergmann", "email": "arnd@arndb.de", "time": "Mon Feb 04 14:28:52 2013 -0800" }, "committer": { "name": "Linus Torvalds", "email": "torvalds@linux-foundation.org", "time": "Tue Feb 05 20:38:49 2013 +1100" }, "message": "samples/seccomp: be less stupid about cross compiling\n\nThe seccomp filters are currently built for the build host, not for the\nmachine that they are going to run on, but they are also built for with\nthe -m32 flag if the kernel is built for a 32 bit machine, both of which\nseems rather odd.\n\nIt broke allyesconfig on my machine, which is x86-64, but building for\n32 bit ARM, with this error message:\n\n In file included from /usr/include/stdio.h:28:0,\n from samples/seccomp/bpf-fancy.c:15:\n /usr/include/features.h:324:26: fatal error: bits/predefs.h: No such file or directory\n\nbecause there are no 32 bit libc headers installed on this machine. We\nshould really be building all the samples for the target machine rather\nthan the build host, but since the infrastructure for that appears to be\nmissing right now, let\u0027s be a little bit smarter and not pass the \u0027-m32\u0027\nflag to the HOSTCC when cross- compiling.\n\nSigned-off-by: Arnd Bergmann \u003carnd@arndb.de\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\nCc: Heiko Carstens \u003cheiko.carstens@de.ibm.com\u003e\nCc: James Morris \u003cjames.l.morris@oracle.com\u003e\nAcked-by: Will Drewry \u003cwad@chromium.org\u003e\nSigned-off-by: Andrew Morton \u003cakpm@linux-foundation.org\u003e\nSigned-off-by: Linus Torvalds \u003ctorvalds@linux-foundation.org\u003e\n" }, { "commit": "b25b09ecf98bf6a32f3732281c2db13be6aeb14c", "tree": "4d7d0e8ddbca40c4d125306e2d9ea3df76f617be", "parents": [ "2e4930eb7c8fb20a39dfb5f8a8f80402710dcea8" ], "author": { "name": "Heiko Carstens", "email": "heiko.carstens@de.ibm.com", "time": "Sat Sep 08 10:23:42 2012 +0200" }, "committer": { "name": "James Morris", "email": "james.l.morris@oracle.com", "time": "Wed Sep 12 12:55:31 2012 +1000" }, "message": "samples/seccomp: fix 31 bit build on s390\n\nOn s390 the flag to force 31 builds is -m31 instead of -m32 unlike\non all (?) other architectures.\n\nFixes this compile error:\n\n HOSTCC samples/seccomp/bpf-direct.o\ncc1: error: unrecognized command line option \"-m32\"\nmake[2]: *** [samples/seccomp/bpf-direct.o] Error 1\n\nSigned-off-by: Heiko Carstens \u003cheiko.carstens@de.ibm.com\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: James Morris \u003cjames.l.morris@oracle.com\u003e\n" }, { "commit": "de4bb3b9c788ea5504dfe094e34d831e8395075d", "tree": "e957a72a3c695d442838b2fef184da8ad5acc193", "parents": [ "663728418e3494f8e4a82f5d1b2f23c22d11be35" ], "author": { "name": "Heiko Carstens", "email": "heiko.carstens@de.ibm.com", "time": "Tue Jul 31 16:15:36 2012 +0200" }, "committer": { "name": "James Morris", "email": "james.l.morris@oracle.com", "time": "Fri Aug 03 14:27:40 2012 +1000" }, "message": "samples/seccomp: fix endianness bug in LO_ARG define\n\nThe LO_ARG define needs to consider endianness also for 32 bit builds.\n\nThe \"bpf_fancy\" test case didn\u0027t work on s390 in 32 bit and compat mode\nbecause the LO_ARG define resulted in a BPF program which read the upper\nhalve of the 64 bit system call arguments instead of the lower halves.\n\nSigned-off-by: Heiko Carstens \u003cheiko.carstens@de.ibm.com\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: James Morris \u003cjames.l.morris@oracle.com\u003e\n" }, { "commit": "8aec836acbe56a27080723187f0a0f1686662318", "tree": "73bb17f8b8d8064a8cdff7334f616406e843afdf", "parents": [ "3072928ff2457642af0da745b88f0420b4596c48" ], "author": { "name": "Chad Williamson", "email": "chad@dahc.us", "time": "Wed Jun 27 04:33:40 2012 -0500" }, "committer": { "name": "James Morris", "email": "james.l.morris@oracle.com", "time": "Fri Jun 29 00:54:17 2012 +1000" }, "message": "samples: seccomp: add .gitignore for untracked executables\n\ngit status should be clean following make allmodconfig \u0026\u0026 make. Add\na .gitignore file to the samples/seccomp directory to ignore binaries\nproduced there.\n\nSigned-off-by: Chad Williamson \u003cchad@dahc.us\u003e\nReviewed-By: Will Drewry \u003cwad@chromium.org\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: James Morris \u003cjames.l.morris@oracle.com\u003e\n" }, { "commit": "561381a146a31ff91d7a2370c10871b02ac7343c", "tree": "e98955e4b362fd25fd8f11603804ecb74c7f4208", "parents": [ "389da25f93eea8ff64181ae7e3e87da68acaef2e" ], "author": { "name": "Will Drewry", "email": "wad@chromium.org", "time": "Wed Apr 18 19:50:25 2012 -0500" }, "committer": { "name": "James Morris", "email": "james.l.morris@oracle.com", "time": "Thu Apr 19 13:44:06 2012 +1000" }, "message": "samples/seccomp: fix dependencies on arch macros\n\nThis change fixes the compilation error triggered here for\ni386 allmodconfig in linux-next:\n http://kisskb.ellerman.id.au/kisskb/buildresult/6123842/\n\nLogic attempting to predict the host architecture has been\nremoved from the Makefile. Instead, the bpf-direct sample\nshould now compile on any architecture, but if the architecture\nis not supported, it will compile a minimal main() function.\n\nThis change also ensures the samples are not compiled when\nthere is no seccomp filter support.\n\n(Note, I wasn\u0027t able to reproduce the error locally, but\n the existing approach was clearly flawed. This tweak\n should resolve your issue and avoid other future weirdness.)\n\nReported-by: Paul Gortmaker \u003cpaul.gortmaker@windriver.com\u003e\nSuggested-by: Kees Cook \u003ckeescook@chromium.org\u003e\nSigned-off-by: Will Drewry \u003cwad@chromium.org\u003e\nSigned-off-by: James Morris \u003cjames.l.morris@oracle.com\u003e\n" }, { "commit": "8ac270d1e29f0428228ab2b9a8ae5e1ed4a5cd84", "tree": "6deba4ed83da9ace758004b29d15aa0d2ec875a7", "parents": [ "c6cfbeb4029610c8c330c312dcf4d514cc067554" ], "author": { "name": "Will Drewry", "email": "wad@chromium.org", "time": "Thu Apr 12 16:48:04 2012 -0500" }, "committer": { "name": "James Morris", "email": "james.l.morris@oracle.com", "time": "Sat Apr 14 11:13:22 2012 +1000" }, "message": "Documentation: prctl/seccomp_filter\n\nDocuments how system call filtering using Berkeley Packet\nFilter programs works and how it may be used.\nIncludes an example for x86 and a semi-generic\nexample using a macro-based code generator.\n\nAcked-by: Eric Paris \u003ceparis@redhat.com\u003e\nSigned-off-by: Will Drewry \u003cwad@chromium.org\u003e\nAcked-by: Kees Cook \u003ckeescook@chromium.org\u003e\n\nv18: - added acked by\n - update no new privs numbers\nv17: - remove @compat note and add Pitfalls section for arch checking\n (keescook@chromium.org)\nv16: -\nv15: -\nv14: - rebase/nochanges\nv13: - rebase on to 88ebdda6159ffc15699f204c33feb3e431bf9bdc\nv12: - comment on the ptrace_event use\n - update arch support comment\n - note the behavior of SECCOMP_RET_DATA when there are multiple filters\n (keescook@chromium.org)\n - lots of samples/ clean up incl 64-bit bpf-direct support\n (markus@chromium.org)\n - rebase to linux-next\nv11: - overhaul return value language, updates (keescook@chromium.org)\n - comment on do_exit(SIGSYS)\nv10: - update for SIGSYS\n - update for new seccomp_data layout\n - update for ptrace option use\nv9: - updated bpf-direct.c for SIGILL\nv8: - add PR_SET_NO_NEW_PRIVS to the samples.\nv7: - updated for all the new stuff in v7: TRAP, TRACE\n - only talk about PR_SET_SECCOMP now\n - fixed bad JLE32 check (coreyb@linux.vnet.ibm.com)\n - adds dropper.c: a simple system call disabler\nv6: - tweak the language to note the requirement of\n PR_SET_NO_NEW_PRIVS being called prior to use. (luto@mit.edu)\nv5: - update sample to use system call arguments\n - adds a \"fancy\" example using a macro-based generator\n - cleaned up bpf in the sample\n - update docs to mention arguments\n - fix prctl value (eparis@redhat.com)\n - language cleanup (rdunlap@xenotime.net)\nv4: - update for no_new_privs use\n - minor tweaks\nv3: - call out BPF \u003c-\u003e Berkeley Packet Filter (rdunlap@xenotime.net)\n - document use of tentative always-unprivileged\n - guard sample compilation for i386 and x86_64\nv2: - move code to samples (corbet@lwn.net)\nSigned-off-by: James Morris \u003cjames.l.morris@oracle.com\u003e\n" } ] }