xref: /third_party/ltp/README.kernel_config

---------------------------------
Enable CODE COVERAGE for your Kernel:
---------------------------------
1) Apply kernel-gcov patch(s) against the corresponding Kernel. They are available here:
http://ltp.cvs.sourceforge.net/viewvc/ltp/utils/analysis/gcov-kernel/linux-2.*.*-gcov.patch?view=log,
2) Also enable the following options in your .config file before building the kernel
CONFIG_GCOV_PROFILE=y
CONFIG_GCOV_ALL=y
CONFIG_GCOV_PROC=m
CONFIG_GCOV_HAMMER=y

---------------------------------
Enabling Kernel Configuration to test Containers/Namespaces
---------------------------------
CONFIG_NAMESPACES=y
CONFIG_UTS_NS=y
CONFIG_IPC_NS=y
CONFIG_USER_NS=y
CONFIG_PID_NS=y
CONFIG_NET_NS=y
CONFIG_VETH=y
CONFIG_MACVLAN=y

The IPC namespaces do not automatically enable IPC, so you may
also want to have:

CONFIG_SYSVIPC=y
CONFIG_SYSVIPC_SYSCTL=y
CONFIG_POSIX_MQUEUE=y

---------------------------------
Enabling Kernel Configuration to test Controllers
---------------------------------
CONFIG_CGROUPS=y
CONFIG_CGROUP_DEBUG=y
CONFIG_CGROUP_NS=y
CONFIG_GROUP_SCHED=y
CONFIG_FAIR_GROUP_SCHED=y
CONFIG_RT_GROUP_SCHED=y
CONFIG_CGROUP_SCHED=y
CONFIG_CGROUP_MEM_RES_CTLR=y
CONFIG_LOCKDEP=y
---------------------------------
Enabling Kernel Configuration to test Power Management features
---------------------------------
CONFIG_CPU_FREQ=y
CONFIG_CPU_FREQ_TABLE=y
CONFIG_CPU_FREQ_DEBUG=y
CONFIG_CPU_FREQ_STAT=y
CONFIG_CPU_FREQ_STAT_DETAILS=y
CONFIG_CPU_IDLE=y
CONFIG_CPU_IDLE_GOV_LADDER=y
CONFIG_CPU_IDLE_GOV_MENU=y
CONFIG_SCHED_MC=y
CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y
CONFIG_CPU_FREQ_DEFAULT_GOV_POWERSAVE=y
CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE=y
CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND=y
CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE=y
CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
CONFIG_CPU_FREQ_GOV_POWERSAVE=y
CONFIG_CPU_FREQ_GOV_USERSPACE=y
CONFIG_CPU_FREQ_GOV_ONDEMAND=y
CONFIG_CPU_FREQ_GOV_CONSERVATIVE=y
---------------------------------
Enabling Kernel Configuration to test filecaps security feature
---------------------------------
CONFIG_SECURITY_FILE_CAPABILITIES=y
---------------------------------
Enabling Kernel Configuration to test SELinux security feature
---------------------------------
Your Kernel should have been built with the following options to
test SELinux:

CONFIG_SECURITY=y
CONFIG_SECURITY_NETWORK=y
CONFIG_SECURITY_NETWORK_XFRM=y

CONFIG_SECURITY_DEFAULT_MMAP_MIN_ADDR=0
This has to be set to a positive value if you want to test this check.
Fedora kernels set it to 65536.

CONFIG_SECURITY_SELINUX=y
CONFIG_SECURITY_SELINUX_BOOTPARAM=y
CONFIG_SECURITY_SELINUX_BOOTPARAM_VALUE=1
CONFIG_SECURITY_SELINUX_DEVELOP=y
CONFIG_SECURITY_SELINUX_AVC_STATS=y
CONFIG_SECURITY_SELINUX_CHECKREQPROT_VALUE=1
CONFIG_SECURITY_SELINUX_ENABLE_SECMARK_DEFAULT=y

CONFIG_SECURITY_SELINUX_POLICYDB_VERSION_MAX=y
You don't want this one unless you are running Fedora 3 or 4.
On anything newer, it will cause unnecessary policy expansion.

CONFIG_SECURITY_SMACK=y
CONFIG_SECURITY_SELINUX=y

By default, if you boot with multiple LSMs compiled into the kernel, the
kernel won't boot succesfully - there can be only one (aside from
explicit internal "stacking" e.g. as is done for combining SELinux or
Smack with capabilities).  Unless you use the security= option to select
one at boot.  SELinux and Smack will honor the security= option.
---------------------------------
---------------------------------
Enabling Kernel Configuration to test SMACK security feature
---------------------------------
CONFIG_NETLABEL=y
CONFIG_SECURITY=y
CONFIG_SECURITY_NETWORK=y
CONFIG_SECURITY_SMACK=y
CONFIG_SECURITY_SELINUX should not be set

For more information to build/install/run these tests, look through:
ltp/testcases/kernel/security/smack/README
---------------------------------
---------------------------------
Enablement for Enhancement to kexec/kdump for implementing the following features:
- Backup/restore memory used by the original kernel before/after kexec.
- Save/restore CPU state before/after kexec.
Now, only the i386 architecture is supported. More from the following git logs:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=3ab83521378268044a448113c6aa9a9e245f4d2f,
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=89081d17f7bb81d89fa1aa9b70f821c5cf4d39e9,
---------------------------------
CONFIG_X86_32=y
CONFIG_RELOCATABLE=y
CONFIG_KEXEC=y
CONFIG_CRASH_DUMP=y
CONFIG_PM=y
CONFIG_HIBERNATION=y
CONFIG_KEXEC_JUMP=y
---------------------------------
Enabling HOTPLUG for your Kernels
---------------------------------
CONFIG_HOTPLUG=y
CONFIG_HOTPLUG_CPU=y
CONFIG_ARCH_ENABLE_MEMORY_HOTPLUG=y
CONFIG_ACPI_HOTPLUG_CPU=y
CONFIG_HOTPLUG_PCI_PCIE=y
CONFIG_HOTPLUG_PCI=y
CONFIG_HOTPLUG_PCI_FAKE=y
CONFIG_HOTPLUG_PCI_COMPAQ=y
CONFIG_HOTPLUG_PCI_COMPAQ_NVRAM=y
CONFIG_HOTPLUG_PCI_IBM=y
CONFIG_HOTPLUG_PCI_ACPI=y
CONFIG_HOTPLUG_PCI_ACPI_IBM=y
CONFIG_HOTPLUG_PCI_CPCI=y
CONFIG_HOTPLUG_PCI_CPCI_ZT5550=y
CONFIG_HOTPLUG_PCI_CPCI_GENERIC=y
CONFIG_HOTPLUG_PCI_SHPC=y
---------------------------------
Video For Linux Two API testing Requirements:
You need to have a video device (i.e. webcam, tuner card, etc.) connected to your system and available under /dev/video0. If you don't have any hardware device available, you can still test the "Virtual Video Driver". To compile this you need to compile your kernel with CONFIG_VIDEO_VIVI=m under:
  -> Device Drivers
    -> Multimedia devices
      -> Video For Linux
        -> Video capture adapters
         -> Virtual Video Driver

The tests also require CUnit Framework to be installed before compiling the tests. Download & Install the same from:
http://sourceforge.net/projects/cunit/
---------------------------------
---------------------------------
Native language support (nls) testsuite requirements
----------------------------------------------------
CONFIG_NLS=m
CONFIG_BLOCK=y
---------------------------------
Enabling Controller area network (CAN) protocol support for your Kernels
---------------------------------
CONFIG_NET=y
CONFIG_CAN=m
CONFIG_CAN_RAW=m
CONFIG_CAN_BCM=m
# CAN Device Drivers
CONFIG_CAN_VCAN=m
---------------------------------
Enabling Fault Injection Support for your kernel (version 2.6.29).
Please check with the original kernel for the fault injection
types it supports. Following supports will be available:

/sys/kernel/debug/fail_io_timeout/interval
/sys/kernel/debug/fail_io_timeout/probability
/sys/kernel/debug/fail_io_timeout/reject-end
/sys/kernel/debug/fail_io_timeout/reject-start
/sys/kernel/debug/fail_io_timeout/require-end
/sys/kernel/debug/fail_io_timeout/require-start
/sys/kernel/debug/fail_io_timeout/space
/sys/kernel/debug/fail_io_timeout/stacktrace-depth
/sys/kernel/debug/fail_io_timeout/task-filter
/sys/kernel/debug/fail_io_timeout/times
/sys/kernel/debug/fail_io_timeout/verbose

/sys/kernel/debug/fail_make_request/interval
/sys/kernel/debug/fail_make_request/probability
/sys/kernel/debug/fail_make_request/reject-end
/sys/kernel/debug/fail_make_request/reject-start
/sys/kernel/debug/fail_make_request/require-end
/sys/kernel/debug/fail_make_request/require-start
/sys/kernel/debug/fail_make_request/space
/sys/kernel/debug/fail_make_request/stacktrace-depth
/sys/kernel/debug/fail_make_request/task-filter
/sys/kernel/debug/fail_make_request/times
/sys/kernel/debug/fail_make_request/verbose

/sys/kernel/debug/fail_page_alloc/ignore-gfp-highmem
/sys/kernel/debug/fail_page_alloc/ignore-gfp-wait
/sys/kernel/debug/fail_page_alloc/interval
/sys/kernel/debug/fail_page_alloc/min-order
/sys/kernel/debug/fail_page_alloc/probability
/sys/kernel/debug/fail_page_alloc/reject-end
/sys/kernel/debug/fail_page_alloc/reject-start
/sys/kernel/debug/fail_page_alloc/require-end
/sys/kernel/debug/fail_page_alloc/require-start
/sys/kernel/debug/fail_page_alloc/space
/sys/kernel/debug/fail_page_alloc/stacktrace-depth
/sys/kernel/debug/fail_page_alloc/task-filter
/sys/kernel/debug/fail_page_alloc/times
/sys/kernel/debug/fail_page_alloc/verbose

/sys/kernel/debug/failslab/ignore-gfp-wait
/sys/kernel/debug/failslab/interval
/sys/kernel/debug/failslab/probability
/sys/kernel/debug/failslab/reject-end
/sys/kernel/debug/failslab/reject-start
/sys/kernel/debug/failslab/require-end
/sys/kernel/debug/failslab/require-start
/sys/kernel/debug/failslab/space
/sys/kernel/debug/failslab/stacktrace-depth
/sys/kernel/debug/failslab/task-filter
/sys/kernel/debug/failslab/times
/sys/kernel/debug/failslab/verbose

when the below kernel config options are set:

CONFIG_FAULT_INJECTION=y
CONFIG_DEBUG_KERNEL=y
CONFIG_FAILSLAB=y (Fault-injection capability for kmalloc)
(CONFIG_SLAB=y || CONFIG_SLUB=y) if CONFIG_FAILSLAB=y
CONFIG_FAIL_PAGE_ALLOC=y (Fault-injection capabilitiy for alloc_pages())
CONFIG_FAIL_MAKE_REQUEST=y (Fault-injection capability for disk IO)
CONFIG_BLOCK=y if CONFIG_FAIL_MAKE_REQUEST=y
CONFIG_FAIL_IO_TIMEOUT=y (Faul-injection capability for faking disk interrupts)
CONFIG_BLOCK=y if CONFIG_FAIL_IO_TIMEOUT=y
CONFIG_FAULT_INJECTION_DEBUG_FS=y (Debugfs entries for fault-injection capabilities)
(CONFIG_SYSFS=y && CONFIG_DEBUG_FS=y) if CONFIG_FAULT_INJECTION_DEBUG_FS=y
CONFIG_FAULT_INJECTION_STACKTRACE_FILTER=y (stacktrace filter for fault-injection capabilities)
(CONFIG_FAULT_INJECTION_DEBUG_FS=y && CONFIG_STACKTRACE_SUPPORT=y && !CONFIG_X86_64) if
	CONFIG_FAULT_INJECTION_STACKTRACE_FILTER=y

For more information on Fault injection, please refer to:
linux-2.6/Documentation/fault-injection/fault-injection.txt,

You should also have made the following entries in your /etc/fstab file
once the kernel is booted with the above CONFIG options set:

debugfs		/sys/kernel/debug/		debugfs

# How the Kernel Fault Injection works for LTP ?

1) Build Kernel with all the above possible kernel CONFIG Options,
2) Create the above entry in /etc/fstab file,
3) Reboot in the new kernel,
4) Goto LTPROOT. Build and Install LTP as per ltp/INSTALL file,
5) Choose your own test(or default) to run with fault injection as follows:
	./runltp -f <command_file> -F <LOOPS>,<FAULT_PROBABILITY>

The algorithm functions like:
loop (for each testcase)
begin
	execute_testcase(inside_stable_kernel)
	begin
		insert_fault_into_kernel()
		loop X Times
		begin
			execute_testcase(inside_fault_kernel)
		end
		restore_kernel_to_normal()
	end
end

# External TODOs:
1) Add or modify testcases when relevant kernel functionality changes.
---------------------------------

---------------------------------
Enabling Kernel Configuration to test ext4 new features
---------------------------------
CONFIG_EXT4_FS=y
CONFIG_EXT4DEV_COMPAT=y
CONFIG_EXT4_FS_XATTR=y
CONFIG_EXT4_FS_POSIX_ACL=y
CONFIG_EXT4_FS_SECURITY=y

Beside that, the following packages are necessary.
  e2fsprogs
  e2fsprogs-devel
  e2fsprogs-libs
And the version of packages must be 1.41.4 or above.

For more information to build/install/run these tests, look through:
ltp/testcases/kernel/fs/ext4-new-features/README
---------------------------------