Wrapper for Syzkaller reproducers
PATCH v2 1/1 Wrapper for Syzkaller reproducers
Allows one to run the Syzkaller reproducers as part of the LTP.
---
V2:
* Will now refuse to compile on anything other than x86_64, because I haven't tested them on anything else.
* Hopefully fixed problem with submodule not being found because I didn't include a commit reference.
* Added explicity 'no' value if --with-syzkaller-repros is not present
* Reduced the default timeout in the test to 20 seconds. I found this allows me to complete testing in reasonable time.
.gitmodules | 4 + configure.ac | 12 ++ include/mk/features.mk.in | 2 + runtest/.gitignore | 1 + testcases/kernel/Makefile | 1 + testcases/kernel/syzkaller-repros/.gitignore | 1 + testcases/kernel/syzkaller-repros/Makefile | 105 +++++++++++++++ testcases/kernel/syzkaller-repros/README.md | 45 +++++++ testcases/kernel/syzkaller-repros/syzwrap.c | 134 +++++++++++++++++++ testcases/linux-arts | 1 + 10 files changed, 306 insertions(+) create mode 100644 runtest/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/Makefile create mode 100644 testcases/kernel/syzkaller-repros/README.md create mode 100644 testcases/kernel/syzkaller-repros/syzwrap.c create mode 160000 testcases/linux-arts
.gitmodules +4 -0
@@ -1,3 +1,7 @@ [submodule "testcases/kernel/mce-test"] path = testcases/kernel/mce-test url = git://git.kernel.org/pub/scm/linux/kernel/git/gong.chen/mce-test.git +[submodule "testcases/linux-arts"] + path = testcases/linux-arts + url = https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-arts.git + shallow = true
configure.ac +12 -0
@@ -187,6 +187,18 @@ else AC_SUBST([WITH_OPEN_POSIX_TESTSUITE],["no"]) fi +AC_ARG_WITH([syzkaller-repros], + [AC_HELP_STRING([--with-syzkaller-repros], + [compile and install Syzkaller reproducers (default=no)])], + [with_syzkaller_repros=$withval] + [with_syzkaller_repros=no] +) +if test "x$with_syzkaller_repros" = xyes; then + AC_SUBST([WITH_SYZKALLER_REPROS],["yes"]) +else + AC_SUBST([WITH_SYZKALLER_REPROS],["no"]) +fi + # TODO: testcases/realtime requires bash and python. AC_ARG_WITH([realtime-testsuite], [AC_HELP_STRING([--with-realtime-testsuite],
include/mk/features.mk.in +2 -0
@@ -47,3 +47,5 @@ WITH_REALTIME_TESTSUITE := no else WITH_REALTIME_TESTSUITE := @WITH_REALTIME_TESTSUITE@ endif + +WITH_SYZKALLER_REPROS := @WITH_SYZKALLER_REPROS@
runtest/.gitignore new file +1 -0
@@ -0,0 +1 @@ +syzkaller*
testcases/kernel/Makefile +1 -0
@@ -53,6 +53,7 @@ SUBDIRS += connectors \ sched \ security \ sound \ + syzkaller-repros \ tracing \ uevents \
testcases/kernel/syzkaller-repros/.gitignore new file +1 -0
@@ -0,0 +1 @@ +syzwrap
testcases/kernel/syzkaller-repros/Makefile new file +105 -0
105 lines not on this page · raw message
testcases/kernel/syzkaller-repros/README.md new file +45 -0
45 lines not on this page · raw message
testcases/kernel/syzkaller-repros/syzwrap.c new file +134 -0
134 lines not on this page · raw message
testcases/linux-arts new file +1 -0
@@ -0,0 +1 @@ +Subproject commit 07759b820a9cbf01333d861d8eb2613b20d1ede4
Signature · 1 line
2.23.0
PATCH v2 1/1 Wrapper for Syzkaller reproducers
Quoted 11 lines · Richard Palethorpe
Allows one to run the Syzkaller reproducers as part of the LTP.
Signed-off-by Richard Palethorpe <[email protected]>---
V2:
* Will now refuse to compile on anything other than
x86_64, because I haven't tested them on anything else.
Quoted 7 lines
* Hopefully fixed problem with submodule not being found because I didn't include a commit reference.
* Added explicity 'no' value if --with-syzkaller-repros is not present
* Reduced the default timeout in the test to 20 seconds. I found this allows me to complete testing in reasonable time.
What's the status of this? This wasn't merged yet, right? I don't see this here:
https://github.com/linux-test-project/ltp/tree/master/testcases/kernel
People are asking how these tests can be run:
https://github.com/dvyukov/syzkaller-repros/pull/1#issuecomment-583996369
Quoted 405 lines
.gitmodules | 4 + configure.ac | 12 ++ include/mk/features.mk.in | 2 + runtest/.gitignore | 1 + testcases/kernel/Makefile | 1 + testcases/kernel/syzkaller-repros/.gitignore | 1 + testcases/kernel/syzkaller-repros/Makefile | 105 +++++++++++++++ testcases/kernel/syzkaller-repros/README.md | 45 +++++++ testcases/kernel/syzkaller-repros/syzwrap.c | 134 +++++++++++++++++++ testcases/linux-arts | 1 + 10 files changed, 306 insertions(+) create mode 100644 runtest/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/Makefile create mode 100644 testcases/kernel/syzkaller-repros/README.md create mode 100644 testcases/kernel/syzkaller-repros/syzwrap.c create mode 160000 testcases/linux-arts.gitmodules +4
-0@@ -1,3 +1,7 @@ [submodule "testcases/kernel/mce-test"] path = testcases/kernel/mce-test url = git://git.kernel.org/pub/scm/linux/kernel/git/gong.chen/mce-test.git +[submodule "testcases/linux-arts"] + path = testcases/linux-arts + url = https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-arts.git + shallow = trueconfigure.ac +12
-0@@ -187,6 +187,18 @@ else AC_SUBST([WITH_OPEN_POSIX_TESTSUITE],["no"]) fi +AC_ARG_WITH([syzkaller-repros], + [AC_HELP_STRING([--with-syzkaller-repros], + [compile and install Syzkaller reproducers (default=no)])], + [with_syzkaller_repros=$withval] + [with_syzkaller_repros=no] +) +if test "x$with_syzkaller_repros" = xyes; then + AC_SUBST([WITH_SYZKALLER_REPROS],["yes"]) +else + AC_SUBST([WITH_SYZKALLER_REPROS],["no"]) +fi + # TODO: testcases/realtime requires bash and python. AC_ARG_WITH([realtime-testsuite], [AC_HELP_STRING([--with-realtime-testsuite],include/mk/features.mk.in +2
-0@@ -47,3 +47,5 @@ WITH_REALTIME_TESTSUITE := no else WITH_REALTIME_TESTSUITE := @WITH_REALTIME_TESTSUITE@ endif + +WITH_SYZKALLER_REPROS := @WITH_SYZKALLER_REPROS@runtest/.gitignore new file +1
-0@@ -0,0 +1 @@ +syzkaller*testcases/kernel/Makefile +1
-0@@ -53,6 +53,7 @@ SUBDIRS += connectors \ sched \ security \ sound \ + syzkaller-repros \ tracing \ uevents \testcases/kernel/syzkaller-repros/.gitignore new file +1
-0@@ -0,0 +1 @@ +syzwraptestcases/kernel/syzkaller-repros/Makefile new file +105
-0@@ -0,0 +1,105 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# Copyright (c) 2019 Linux Test Project + +top_srcdir ?= ../../.. + +include $(top_srcdir)/include/mk/testcases.mk + +CFLAGS += -D_GNU_SOURCE + +ifeq ($(WITH_SYZKALLER_REPROS),yes) + +# This is mainly due to the -m32 flag, but there could be other problems. +ifneq ($(HOST_CPU), x86_64)) +$(error "We currently only support building the Syzkaller reproducers on x86_64") +endif + +# The number of reproducers in each runtest file +SYZKALLER_RUNFILES_SIZE ?= 100 + +# Extra arguments to pass to syzwrap. Uncomment the below to add some +# sandboxing. +# SYZWRAP_ARGS ?= -s + +# Location where reproducers are installed +SYZKALLER_INSTALL_DIR ?= $(abspath $(DESTDIR)/$(prefix)/testcases/bin) + +# If the reproducers directory is missing then we automatically clone the repo. +# We then have to call make recursively to revaluate the targets +SYZKALLER_REPROS_DIR ?= $(abs_top_srcdir)/testcases/linux-arts/syzkaller-repros/linux +$(SYZKALLER_REPROS_DIR): + git submodule update --init $(abs_top_srcdir)/testcases/linux-arts + $(MAKE) syzkaller_runfiles + +SYZKALLER_REPROS_SRCS = $(wildcard $(SYZKALLER_REPROS_DIR)/*.c) + +# Some useful compiler flags for the LTP will cause problems with the +# syzkaller repros so the repros have seperate flags +SYZKALLER_CFLAGS ?= -pthread +SYZKALLER_REPROS = $(subst $(abs_top_srcdir),$(abs_top_builddir),$(SYZKALLER_REPROS_SRCS:.c=)) +$(SYZKALLER_REPROS): %: %.c + -@if grep -q "__NR_mmap2" $^; then \ + M32="-m32"; \ + fi; \ + $(CC) $(SYZKALLER_CFLAGS) $$M32 $(SYZKALLER_LDFLAGS) $^ -o $@; \ + echo $(CC) $(SYZKALLER_CFLAGS) $$M32 $(SYZKALLER_LDFLAGS) $^ -o $@; + +# Generate the names of the runtest files. This uses Shell arithmetic to +# calculate how many runtest files there will be. +define SYZKALLER_RUNFILES != + n=$(words $(SYZKALLER_REPROS)); + m=$(SYZKALLER_RUNFILES_SIZE); + i=$$(( $$n / $$m + ($$n % $$m > 0) )); + while test $$i -gt 0; + do + echo $(top_srcdir)/runtest/syzkaller$$i; + i=$$(($$i - 1)); + done +endef + +# Get the index part of a runtest files name +syz_n = $(subst $(top_srcdir)/runtest/syzkaller,,$(1)) +syz_m = $(SYZKALLER_RUNFILES_SIZE) +# Gives the index of the first reproducer in a runtest file +syz_i = $(shell echo $$((($(call syz_n,$(1)) - 1) * $(2) + 1))) +# Gives the index of the last reproducer in a runtest file +syz_j = $(shell echo $$(( $(call syz_i,$(1),$(2)) + $(2) - 1 ))) +# Gvien a runtest file name, get the reproducers it should contain +syz_wordlist = $(wordlist $(call syz_i,$(1),$(syz_m)),$(call syz_j,$(1),$(syz_m)),$(SYZKALLER_REPROS)) + +define syz_runfile_line +$(notdir $(exe)) syzwrap $(SYZWRAP_ARGS) -d $(SYZKALLER_INSTALL_DIR) -n $(notdir $(exe)) + +endef + +# Generate the runtest files based on the reproducer names and batch size. +$(SYZKALLER_RUNFILES): $(SYZKALLER_REPROS) + @echo "Writing $@" + $(file >$@) + $(foreach exe,$(call syz_wordlist,$@),$(file >>$@,$(syz_runfile_line))) + +.PHONY: syzkaller_runfiles +syzkaller_runfiles: $(SYZKALLER_RUNFILES) | $(SYZKALLER_REPROS_DIR) + +all: $(SYZKALLER_RUNFILES) | $(SYZKALLER_REPROS_DIR) + +# There are too many reproducers to pass all at once to rm, so we just pass +# one at a time +syzkaller_clean: + $(foreach f, $(SYZKALLER_REPROS), $(RM) $(f)) +CLEAN_DEPS += syzkaller_clean +CLEAN_TARGETS += $(SYZKALLER_RUNFILES) + +INSTALL_MODE ?= 0775 + +# For some reason part of the path is missing if we just try to install these +# by adding them to INSTALL_FILES +SYZKALLER_REPROS_INSTALLED := $(subst $(SYZKALLER_REPROS_DIR),$(SYZKALLER_INSTALL_DIR),$(SYZKALLER_REPROS)) +$(SYZKALLER_REPROS_INSTALLED): $(SYZKALLER_INSTALL_DIR)/%: $(SYZKALLER_REPROS_DIR)/% + install -m $(INSTALL_MODE) -T $< $@ + +install: $(SYZKALLER_REPROS_INSTALLED) + +endif + +include $(top_srcdir)/include/mk/generic_leaf_target.mktestcases/kernel/syzkaller-repros/README.md new file +45
-0@@ -0,0 +1,45 @@ +# LTP wrapper for Syzkaller reproducers + +This allows you to run the autogenerated C bug reproducers from the Syzkaller +fuzzer within the LTP framework. Meaning that you may use an existing test +runner compatible with the LTP (with some constraints, see below). + +## Instructions + +1. Run `ltp/configure` with `--with-syzkaller-repros`. +2. Build and install the LTP as normal. +3. Run one or more of syzkallerN runtest files where N is a number. + +Make will automatically download the reproducers into `testcases/linux-arts` +using git-submodule if necessary. + +By default each runtest file contains 100 reproducers. You may change this by +overriding `SYZKALLER_RUNFILES_SIZE`. + +Extra parameters can be sent to syzwrap using `SYZWRAP_ARGS`. See `syzwrap +-h`. + +## Kernel Requirements + +It is strongly recommended that you use KASAN and other debugging kernel +features. See the Syzkaller documentation for the configuration you should +use. + +## Test Runner Requirements + +Unlike most LTP tests these reproducers can leave your system in a broken +state even if no bug is triggered. + +You will need to: + +A) Reboot the SUT +B) Reset at least the root filesystem to a known good state + +Every time syzwrap fails. + +If syzwrap fails with TBROK or fails to run at all, then you probably need to +reset the system and rerun that test. If a test fails with TFAIL, you may also +want to run it once again with a clean state. + +It might be the case that some reproducers write to random devices or do other +things which can effect the outside world.testcases/kernel/syzkaller-repros/syzwrap.c new file +134
-0@@ -0,0 +1,134 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* +* Copyright (c) 2019 Richard Palethorpe <[email protected]> +* +* Run a single reproducer generated by the Syzkaller fuzzer. +*/ + +#include <sys/types.h> +#include <sys/wait.h> +#include <sys/prctl.h> +#include <sched.h> +#include <signal.h> +#include <stdio.h> +#include <pwd.h> + +#include "tst_test.h" +#include "tst_taint.h" +#include "tst_safe_stdio.h" + +#define SANDBOX_HELP "\n"\ + "-s\t Add some sandboxing around the reproducer. This will prevent some\n"\ + "\t reproducers from creating network devices and thus prevent them from\n"\ + "\t working. However it will also prevent some reproducers from trashing\n"\ + "\t the system using root privileges. Note that you may generate the\n"\ + "\t reproducers with various types of sandboxing built in using\n"\ + "\t syz-reprolist" + +static char *dir; +static char *name; +static char *path; + +static char *sandbox; + +static struct tst_option options[] = { + {"d:" , &dir,"\n-d PATH\t Mandatory directory containing reproducers" }, + {"n:" , &name,"-n NAME\t Mandatory executable name of reproducer" }, + {"s" , &sandbox,SANDBOX_HELP }, + {NULL ,NULL ,NULL } +}; + +static void become_nobody (void ) +{ +struct passwd *pw; +int gid, uid; + +setgroups (0 ,NULL ); + + pw =getpwnam ("nobody" ); +if (pw) { + gid = pw->pw_gid; + uid = pw->pw_uid; + }else { + gid =65534 ; + uid =65534 ; + } + +SAFE_SETREGID (gid, gid); +SAFE_SETREUID (uid, uid); +} + +static void setup (void ) +{ +tst_taint_init (TST_TAINT_W |TST_TAINT_D |TST_TAINT_L ); + +if (!dir) +tst_brk (TBROK ,"No reproducer directory specified" ); + +if (!name) +tst_brk (TBROK ,"No reproducer name specified" ); + +tst_res (TINFO ,"https://syzkaller.appspot.com/bug?id=%s" , name); + +SAFE_ASPRINTF (&path,"%s/%s" , dir, name); +tst_res (TINFO ,"%s" , path); +} + +static void run (void ) +{ +unsigned int backoff =100 ; +int rem, status, sent_kill =0 ; +float exec_time_start = (float )tst_timeout_remaining (); +int pid; + +if (sandbox) +SAFE_UNSHARE (CLONE_NEWPID ); + + pid =SAFE_FORK (); +if (!pid) { +if (sandbox) { +SAFE_UNSHARE (CLONE_NEWNET ); +become_nobody (); + } + +if (prctl (PR_SET_DUMPABLE ,1 ,0 ,0 ,0 )) { +tst_res (TWARN |TERRNO , +"Failed to set dumpable; won't be able to open /proc/self/*" ); + } + +execl (path, name,NULL ); +tst_brk (TBROK |TERRNO ,"Failed to exec reproducer" ); + } + +while (!waitpid (pid, &status,WNOHANG )) { + rem =tst_timeout_remaining (); + +if (!sent_kill && rem / exec_time_start <0.5 ) { +tst_res (TINFO ,"Timeout; killing reproducer" ); + +TEST (kill (pid,SIGKILL )); +if (TST_RET ==-1 ) +tst_res (TWARN |TTERRNO ,"kill() failed" ); +else + sent_kill =1 ; + } + +usleep (backoff); + backoff =MIN (2 * backoff,1000000 ); + } + +if (tst_taint_check ()) { +tst_res (TFAIL ,"Kernel is tainted" ); + }else { +tst_res (TPASS ,"Kernel is not tainted" ); + } +} + +static struct tst_test test = { + .setup = setup, + .test_all = run, + .options = options, + .needs_tmpdir =1 , + .forks_child =1 , + .timeout =20 +};testcases/linux-arts new file +1
-1@@ -0,0 +1 @@ +Subproject commit 07759b820a9cbf01333d861d8eb2613b20d1ede4 --2.23.0
--
You received this message because you are subscribed to the Google Groups "syzkaller" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/syzkaller/20191204103120.29440-1-rpalethorpe%40suse.com.
PATCH v2 1/1 Wrapper for Syzkaller reproducers
Hello Dmitry,
Quoted 15 lines · Dmitry Vyukov via ltp
Quoted 14 lines · Richard Palethorpe
Allows one to run the Syzkaller reproducers as part of the LTP.
Signed-off-by Richard Palethorpe <[email protected]>---
V2:
* Will now refuse to compile on anything other than
x86_64, because I haven't tested them on anything else.* Hopefully fixed problem with submodule not being found because I didn't include a commit reference.
Quoted 8 lines
Quoted 4 lines
* Added explicity 'no' value if --with-syzkaller-repros is not present
* Reduced the default timeout in the test to 20 seconds. I found this allows me to complete testing in reasonable time.
What's the status of this? This wasn't merged yet, right? I don't see this here:
https://github.com/linux-test-project/ltp/tree/master/testcases/kernel
Yeah, it fell to one side due to the LTP release, SUSE Hackweek and lack of feedback. Hopefully we can get it merged soon however.
Quoted 3 lines
People are asking how these tests can be run:
https://github.com/dvyukov/syzkaller-repros/pull/1#issuecomment-583996369
I responded with some links for using it.
Quoted 403 lines
Quoted 403 lines
.gitmodules | 4 + configure.ac | 12 ++ include/mk/features.mk.in | 2 + runtest/.gitignore | 1 + testcases/kernel/Makefile | 1 + testcases/kernel/syzkaller-repros/.gitignore | 1 + testcases/kernel/syzkaller-repros/Makefile | 105 +++++++++++++++ testcases/kernel/syzkaller-repros/README.md | 45 +++++++ testcases/kernel/syzkaller-repros/syzwrap.c | 134 +++++++++++++++++++ testcases/linux-arts | 1 + 10 files changed, 306 insertions(+) create mode 100644 runtest/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/.gitignore create mode 100644 testcases/kernel/syzkaller-repros/Makefile create mode 100644 testcases/kernel/syzkaller-repros/README.md create mode 100644 testcases/kernel/syzkaller-repros/syzwrap.c create mode 160000 testcases/linux-arts.gitmodules +4
-0@@ -1,3 +1,7 @@ [submodule "testcases/kernel/mce-test"] path = testcases/kernel/mce-test url = git://git.kernel.org/pub/scm/linux/kernel/git/gong.chen/mce-test.git +[submodule "testcases/linux-arts"] + path = testcases/linux-arts + url = https://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-arts.git + shallow = trueconfigure.ac +12
-0@@ -187,6 +187,18 @@ else AC_SUBST([WITH_OPEN_POSIX_TESTSUITE],["no"]) fi +AC_ARG_WITH([syzkaller-repros], + [AC_HELP_STRING([--with-syzkaller-repros], + [compile and install Syzkaller reproducers (default=no)])], + [with_syzkaller_repros=$withval] + [with_syzkaller_repros=no] +) +if test "x$with_syzkaller_repros" = xyes; then + AC_SUBST([WITH_SYZKALLER_REPROS],["yes"]) +else + AC_SUBST([WITH_SYZKALLER_REPROS],["no"]) +fi + # TODO: testcases/realtime requires bash and python. AC_ARG_WITH([realtime-testsuite], [AC_HELP_STRING([--with-realtime-testsuite],include/mk/features.mk.in +2
-0@@ -47,3 +47,5 @@ WITH_REALTIME_TESTSUITE := no else WITH_REALTIME_TESTSUITE := @WITH_REALTIME_TESTSUITE@ endif + +WITH_SYZKALLER_REPROS := @WITH_SYZKALLER_REPROS@runtest/.gitignore new file +1
-0@@ -0,0 +1 @@ +syzkaller*testcases/kernel/Makefile +1
-0@@ -53,6 +53,7 @@ SUBDIRS += connectors \ sched \ security \ sound \ + syzkaller-repros \ tracing \ uevents \testcases/kernel/syzkaller-repros/.gitignore new file +1
-0@@ -0,0 +1 @@ +syzwraptestcases/kernel/syzkaller-repros/Makefile new file +105
-0@@ -0,0 +1,105 @@ +# SPDX-License-Identifier: GPL-2.0-or-later +# Copyright (c) 2019 Linux Test Project + +top_srcdir ?= ../../.. + +include $(top_srcdir)/include/mk/testcases.mk + +CFLAGS += -D_GNU_SOURCE + +ifeq ($(WITH_SYZKALLER_REPROS),yes) + +# This is mainly due to the -m32 flag, but there could be other problems. +ifneq ($(HOST_CPU), x86_64)) +$(error "We currently only support building the Syzkaller reproducers on x86_64") +endif + +# The number of reproducers in each runtest file +SYZKALLER_RUNFILES_SIZE ?= 100 + +# Extra arguments to pass to syzwrap. Uncomment the below to add some +# sandboxing. +# SYZWRAP_ARGS ?= -s + +# Location where reproducers are installed +SYZKALLER_INSTALL_DIR ?= $(abspath $(DESTDIR)/$(prefix)/testcases/bin) + +# If the reproducers directory is missing then we automatically clone the repo. +# We then have to call make recursively to revaluate the targets +SYZKALLER_REPROS_DIR ?= $(abs_top_srcdir)/testcases/linux-arts/syzkaller-repros/linux +$(SYZKALLER_REPROS_DIR): + git submodule update --init $(abs_top_srcdir)/testcases/linux-arts + $(MAKE) syzkaller_runfiles + +SYZKALLER_REPROS_SRCS = $(wildcard $(SYZKALLER_REPROS_DIR)/*.c) + +# Some useful compiler flags for the LTP will cause problems with the +# syzkaller repros so the repros have seperate flags +SYZKALLER_CFLAGS ?= -pthread +SYZKALLER_REPROS = $(subst $(abs_top_srcdir),$(abs_top_builddir),$(SYZKALLER_REPROS_SRCS:.c=)) +$(SYZKALLER_REPROS): %: %.c + -@if grep -q "__NR_mmap2" $^; then \ + M32="-m32"; \ + fi; \ + $(CC) $(SYZKALLER_CFLAGS) $$M32 $(SYZKALLER_LDFLAGS) $^ -o $@; \ + echo $(CC) $(SYZKALLER_CFLAGS) $$M32 $(SYZKALLER_LDFLAGS) $^ -o $@; + +# Generate the names of the runtest files. This uses Shell arithmetic to +# calculate how many runtest files there will be. +define SYZKALLER_RUNFILES != + n=$(words $(SYZKALLER_REPROS)); + m=$(SYZKALLER_RUNFILES_SIZE); + i=$$(( $$n / $$m + ($$n % $$m > 0) )); + while test $$i -gt 0; + do + echo $(top_srcdir)/runtest/syzkaller$$i; + i=$$(($$i - 1)); + done +endef + +# Get the index part of a runtest files name +syz_n = $(subst $(top_srcdir)/runtest/syzkaller,,$(1)) +syz_m = $(SYZKALLER_RUNFILES_SIZE) +# Gives the index of the first reproducer in a runtest file +syz_i = $(shell echo $$((($(call syz_n,$(1)) - 1) * $(2) + 1))) +# Gives the index of the last reproducer in a runtest file +syz_j = $(shell echo $$(( $(call syz_i,$(1),$(2)) + $(2) - 1 ))) +# Gvien a runtest file name, get the reproducers it should contain +syz_wordlist = $(wordlist $(call syz_i,$(1),$(syz_m)),$(call syz_j,$(1),$(syz_m)),$(SYZKALLER_REPROS)) + +define syz_runfile_line +$(notdir $(exe)) syzwrap $(SYZWRAP_ARGS) -d $(SYZKALLER_INSTALL_DIR) -n $(notdir $(exe)) + +endef + +# Generate the runtest files based on the reproducer names and batch size. +$(SYZKALLER_RUNFILES): $(SYZKALLER_REPROS) + @echo "Writing $@" + $(file >$@) + $(foreach exe,$(call syz_wordlist,$@),$(file >>$@,$(syz_runfile_line))) + +.PHONY: syzkaller_runfiles +syzkaller_runfiles: $(SYZKALLER_RUNFILES) | $(SYZKALLER_REPROS_DIR) + +all: $(SYZKALLER_RUNFILES) | $(SYZKALLER_REPROS_DIR) + +# There are too many reproducers to pass all at once to rm, so we just pass +# one at a time +syzkaller_clean: + $(foreach f, $(SYZKALLER_REPROS), $(RM) $(f)) +CLEAN_DEPS += syzkaller_clean +CLEAN_TARGETS += $(SYZKALLER_RUNFILES) + +INSTALL_MODE ?= 0775 + +# For some reason part of the path is missing if we just try to install these +# by adding them to INSTALL_FILES +SYZKALLER_REPROS_INSTALLED := $(subst $(SYZKALLER_REPROS_DIR),$(SYZKALLER_INSTALL_DIR),$(SYZKALLER_REPROS)) +$(SYZKALLER_REPROS_INSTALLED): $(SYZKALLER_INSTALL_DIR)/%: $(SYZKALLER_REPROS_DIR)/% + install -m $(INSTALL_MODE) -T $< $@ + +install: $(SYZKALLER_REPROS_INSTALLED) + +endif + +include $(top_srcdir)/include/mk/generic_leaf_target.mktestcases/kernel/syzkaller-repros/README.md new file +45
-0@@ -0,0 +1,45 @@ +# LTP wrapper for Syzkaller reproducers + +This allows you to run the autogenerated C bug reproducers from the Syzkaller +fuzzer within the LTP framework. Meaning that you may use an existing test +runner compatible with the LTP (with some constraints, see below). + +## Instructions + +1. Run `ltp/configure` with `--with-syzkaller-repros`. +2. Build and install the LTP as normal. +3. Run one or more of syzkallerN runtest files where N is a number. + +Make will automatically download the reproducers into `testcases/linux-arts` +using git-submodule if necessary. + +By default each runtest file contains 100 reproducers. You may change this by +overriding `SYZKALLER_RUNFILES_SIZE`. + +Extra parameters can be sent to syzwrap using `SYZWRAP_ARGS`. See `syzwrap +-h`. + +## Kernel Requirements + +It is strongly recommended that you use KASAN and other debugging kernel +features. See the Syzkaller documentation for the configuration you should +use. + +## Test Runner Requirements + +Unlike most LTP tests these reproducers can leave your system in a broken +state even if no bug is triggered. + +You will need to: + +A) Reboot the SUT +B) Reset at least the root filesystem to a known good state + +Every time syzwrap fails. + +If syzwrap fails with TBROK or fails to run at all, then you probably need to +reset the system and rerun that test. If a test fails with TFAIL, you may also +want to run it once again with a clean state. + +It might be the case that some reproducers write to random devices or do other +things which can effect the outside world.testcases/kernel/syzkaller-repros/syzwrap.c new file +134
-0@@ -0,0 +1,134 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* +* Copyright (c) 2019 Richard Palethorpe <[email protected]> +* +* Run a single reproducer generated by the Syzkaller fuzzer. +*/ + +#include <sys/types.h> +#include <sys/wait.h> +#include <sys/prctl.h> +#include <sched.h> +#include <signal.h> +#include <stdio.h> +#include <pwd.h> + +#include "tst_test.h" +#include "tst_taint.h" +#include "tst_safe_stdio.h" + +#define SANDBOX_HELP "\n"\ + "-s\t Add some sandboxing around the reproducer. This will prevent some\n"\ + "\t reproducers from creating network devices and thus prevent them from\n"\ + "\t working. However it will also prevent some reproducers from trashing\n"\ + "\t the system using root privileges. Note that you may generate the\n"\ + "\t reproducers with various types of sandboxing built in using\n"\ + "\t syz-reprolist" + +static char *dir; +static char *name; +static char *path; + +static char *sandbox; + +static struct tst_option options[] = { + {"d:" , &dir,"\n-d PATH\t Mandatory directory containing reproducers" }, + {"n:" , &name,"-n NAME\t Mandatory executable name of reproducer" }, + {"s" , &sandbox,SANDBOX_HELP }, + {NULL ,NULL ,NULL } +}; + +static void become_nobody (void ) +{ +struct passwd *pw; +int gid, uid; + +setgroups (0 ,NULL ); + + pw =getpwnam ("nobody" ); +if (pw) { + gid = pw->pw_gid; + uid = pw->pw_uid; + }else { + gid =65534 ; + uid =65534 ; + } + +SAFE_SETREGID (gid, gid); +SAFE_SETREUID (uid, uid); +} + +static void setup (void ) +{ +tst_taint_init (TST_TAINT_W |TST_TAINT_D |TST_TAINT_L ); + +if (!dir) +tst_brk (TBROK ,"No reproducer directory specified" ); + +if (!name) +tst_brk (TBROK ,"No reproducer name specified" ); + +tst_res (TINFO ,"https://syzkaller.appspot.com/bug?id=%s" , name); + +SAFE_ASPRINTF (&path,"%s/%s" , dir, name); +tst_res (TINFO ,"%s" , path); +} + +static void run (void ) +{ +unsigned int backoff =100 ; +int rem, status, sent_kill =0 ; +float exec_time_start = (float )tst_timeout_remaining (); +int pid; + +if (sandbox) +SAFE_UNSHARE (CLONE_NEWPID ); + + pid =SAFE_FORK (); +if (!pid) { +if (sandbox) { +SAFE_UNSHARE (CLONE_NEWNET ); +become_nobody (); + } + +if (prctl (PR_SET_DUMPABLE ,1 ,0 ,0 ,0 )) { +tst_res (TWARN |TERRNO , +"Failed to set dumpable; won't be able to open /proc/self/*" ); + } + +execl (path, name,NULL ); +tst_brk (TBROK |TERRNO ,"Failed to exec reproducer" ); + } + +while (!waitpid (pid, &status,WNOHANG )) { + rem =tst_timeout_remaining (); + +if (!sent_kill && rem / exec_time_start <0.5 ) { +tst_res (TINFO ,"Timeout; killing reproducer" ); + +TEST (kill (pid,SIGKILL )); +if (TST_RET ==-1 ) +tst_res (TWARN |TTERRNO ,"kill() failed" ); +else + sent_kill =1 ; + } + +usleep (backoff); + backoff =MIN (2 * backoff,1000000 ); + } + +if (tst_taint_check ()) { +tst_res (TFAIL ,"Kernel is tainted" ); + }else { +tst_res (TPASS ,"Kernel is not tainted" ); + } +} + +static struct tst_test test = { + .setup = setup, + .test_all = run, + .options = options, + .needs_tmpdir =1 , + .forks_child =1 , + .timeout =20 +};testcases/linux-arts new file +1
-1@@ -0,0 +1 @@ +Subproject commit 07759b820a9cbf01333d861d8eb2613b20d1ede4 --2.23.0
Quoted 4 lines
Quoted 4 lines
--
You received this message because you are subscribed to the Google Groups "syzkaller" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/syzkaller/20191204103120.29440-1-rpalethorpe%40suse.com.
--
Thank you,
Richard.
PATCH v2 1/1 Wrapper for Syzkaller reproducers
Hi!
Quoted 18 lines
Quoted 18 lines
Quoted 17 lines
Allows one to run the Syzkaller reproducers as part of the LTP.
Signed-off-by Richard Palethorpe <[email protected]>---
V2:
* Will now refuse to compile on anything other than
x86_64, because I haven't tested them on anything else.* Hopefully fixed problem with submodule not being found because I didn't include a commit reference.
* Added explicity 'no' value if --with-syzkaller-repros is not present
* Reduced the default timeout in the test to 20 seconds. I found this allows me to complete testing in reasonable time.
Quoted 6 lines
Quoted 3 lines
What's the status of this? This wasn't merged yet, right? I don't see this here:
https://github.com/linux-test-project/ltp/tree/master/testcases/kernelYeah, it fell to one side due to the LTP release, SUSE Hackweek and lack of feedback. Hopefully we can get it merged soon however.
You don't forget about FOSDEM. Unfortunatelly we have managed to accumulate quite a backlog at the start of the year.
However I do plan to return to this once things settle down a bit.
Signature · 2 lines
Cyril Hrubis [email protected]