From: Octavian Purdila <tavi@xxxxxxxxx> The Linux kernel labs documentation is a collection of "labs" for various device driver topics. For each topic there are two parts: a walk-through which explain the basic concepts and a hands-on part which contains a few exercises. This commit also adds the labs infrastructure which allows us to build and test kernel modules in a qemu environment. Signed-off-by: Octavian Purdila <tavi@xxxxxxxxx> Signed-off-by: Daniel Baluta <daniel.baluta@xxxxxxxxx> --- Documentation/labs/conf.py | 1 + Documentation/labs/exercises-summary.hrst | 42 ++++++++ Documentation/labs/exercises.rst | 82 ++++++++++++++++ Documentation/labs/index.rst | 33 +++++++ Documentation/labs/subst.hrst | 4 + Documentation/labs/vm.rst | 157 ++++++++++++++++++++++++++++++ README.rst | 1 + tools/labs/.gitignore | 7 ++ tools/labs/Makefile | 42 ++++++++ tools/labs/qemu/Makefile | 61 ++++++++++++ tools/labs/qemu/create_net.sh | 37 +++++++ tools/labs/qemu/kernel_config.x86 | 79 +++++++++++++++ tools/labs/qemu/prepare-image.sh | 28 ++++++ tools/labs/qemu/qemu.sh | 21 ++++ tools/labs/templates/generate_skels.py | 55 +++++++++++ 15 files changed, 650 insertions(+) create mode 100644 Documentation/labs/conf.py create mode 100644 Documentation/labs/exercises-summary.hrst create mode 100644 Documentation/labs/exercises.rst create mode 100644 Documentation/labs/index.rst create mode 100644 Documentation/labs/subst.hrst create mode 100644 Documentation/labs/vm.rst create mode 120000 README.rst create mode 100644 tools/labs/.gitignore create mode 100644 tools/labs/Makefile create mode 100644 tools/labs/qemu/Makefile create mode 100755 tools/labs/qemu/create_net.sh create mode 100644 tools/labs/qemu/kernel_config.x86 create mode 100755 tools/labs/qemu/prepare-image.sh create mode 100755 tools/labs/qemu/qemu.sh create mode 100755 tools/labs/templates/generate_skels.py diff --git a/Documentation/labs/conf.py b/Documentation/labs/conf.py new file mode 100644 index 0000000..25a53c0 --- /dev/null +++ b/Documentation/labs/conf.py @@ -0,0 +1 @@ +rst_prolog = ".. include:: subst.hrst" diff --git a/Documentation/labs/exercises-summary.hrst b/Documentation/labs/exercises-summary.hrst new file mode 100644 index 0000000..0cddaa2 --- /dev/null +++ b/Documentation/labs/exercises-summary.hrst @@ -0,0 +1,42 @@ + +To solve exercises need to perform these steps: + + * prepare skeletons from templates + * build modules + * copy modules to VM + * start the VM and test the module in the VM. + +The current lab name is |LAB_NAME|. See the exercises for the task name. + +.. container:: toggle + + .. container:: header + + **See details** + + The skeleton code is generated from full source examples located in + ``tools/labs/templates``. To solve the tasks start by generating + the skeleton code: + + .. parsed-literal:: shell + + tools/labs $ make clean + tools/labs $ LABS=<lab name>/<task name> make skels + + Where task name is defined for each task. Once the skelton drivers are + generated build the source: + + .. code-block:: shell + + tools/labs $ make build + + Then copy the modules and start the VM: + + .. code-block:: shell + + tools/labs $ make copy + tools/labs $ make boot + + The modules are placed in /home/root/skels/|LAB_NAME|/<task_name>. + + Review the `Exercises`_ section for more detailed information. diff --git a/Documentation/labs/exercises.rst b/Documentation/labs/exercises.rst new file mode 100644 index 0000000..7120efa --- /dev/null +++ b/Documentation/labs/exercises.rst @@ -0,0 +1,82 @@ +Exercises +========= + +In order to facilitate learning each topic has a hands-on exercises +section which will contain in-depth, incremental clues on how to solve +one or multiple tasks. To focus on a particular issue most of the +tasks will be performed on existing skeleton drivers. Each skeleton +driver has clearly marked sections that needs to be filled in order to +complete the tasks. + +The skeleton drivers are generated from full source examples located +in tools/labs/templates. To solve tasks you start by generating the +skeleton drivers, running the **skels** target in *tools/labs*. To +keep the workspace clean it is recommended to generate the skeletons +for one lab only and clean the workspace before start working on a new +lab. Labs can be selecting by using the **LABS** variable: + +.. code-block:: shell + + tools/labs $ make clean + tools/labs $ LABS=kernel_modules make skels + + tools/labs $ ls skels/kernel_modules/ + 1-2-test-mod 3-error-mod 4-multi-mod 5-oops-mod 6-cmd-mod \ + 7-list-proc 8-kprobes 9-kdb + +You can also uses the same variable to generate skeletons for specific +tasks: + +.. code-block:: shell + + tools/labs $ LABS="kernel_modules/6-cmd-mod kernel_modules/8-kprobes" make skels + + tools/labs$ ls skels/kernel_modules + 6-cmd-mod 8-kprobes + + +For each task you may have multiple steps to perform, usually +incremental. These steps are marked in the code source as well as in +the lab exercises with the keyword *TODO*. If we have multiple steps +to perform they will be prefix by a number, like *TODO1*, *TODO2*, +etc. If no number is used it is assumed to be the one and only +step. If you want to resume a tasks from a certain step, you can using +the **TODO** variable. The following example will generate the +skeleton with the first *TODO* step resolved: + +.. code-block:: shell + + tools/labs $ TODO=2 LABS="kernel_modules/8-kprobes" skels + +Once the skelton drivers are generated you can build them with the +**build** make target: + +.. code-block:: shell + + tools/labs $ make build + echo "# autogenerated, do not edit " > skels/Kbuild + for i in ./kernel_modules/8-kprobes; do echo "obj-m += $i/" >> skels/Kbuild; done + make -C /home/tavi/src/linux M=/home/tavi/src/linux/tools/labs/skels ARCH=x86 modules + make[1]: Entering directory '/home/tavi/src/linux' + CC [M] /home/tavi/src/linux/tools/labs/skels/./kernel_modules/8-kprobes/kprobes.o + Building modules, stage 2. + MODPOST 1 modules + CC /home/tavi/src/linux/tools/labs/skels/./kernel_modules/8-kprobes/kprobes.mod.o + LD [M] /home/tavi/src/linux/tools/labs/skels/./kernel_modules/8-kprobes/kprobes.ko + make[1]: Leaving directory '/home/tavi/src/linux' + + +To copy the drivers to the VM you can use either use ssh or update the +VM image directly using the **copy** target: + +.. code-block:: shell + + tools/labs $ make copy + ... + 'skels/kernel_modules/8-kprobes/kprobes.ko' -> '/tmp/tmp.4UMKcISmQM/home/root/skels/kernel_modules/8-kprobes/kprobes.ko' + +.. attention:: The **copy** target will faili if the VM is + running. This is intentional so that the we avoid corrupting the + filesystem. + + diff --git a/Documentation/labs/index.rst b/Documentation/labs/index.rst new file mode 100644 index 0000000..9e61a8e --- /dev/null +++ b/Documentation/labs/index.rst @@ -0,0 +1,33 @@ +Linux device driver labs +======================== + +This is a collection of "howtos" for various device driver topics. For +each topic there are two parts: + +* a walk-through the topic which contains an overview, the main + abstractions, simple examples and pointers to APIs + +* a hands-on part which contains a few exercises that should be + resolved by the student; to focus on the topic at hand, the student + is presented with a starting coding skeleton and with in-depth tips + on how to solve the exercises + +They are based on the labs of the `Operatings Systems 2 +<http://ocw.cs.pub.ro/courses/so2>`_ course, Computer Science +department, Faculty of Automatic Control and Computers, Politehnica +Univesity of Bucharest. + +You can get the latest version at http://github.com/linux-kernel-labs. + +To get started build the documentation from the sources: + +.. code-block:: c + + cd tools/labs && make docs + +then point your browser at **Documentation/output/labs/index.html**. + +.. toctree:: + + vm.rst + exercises.rst diff --git a/Documentation/labs/subst.hrst b/Documentation/labs/subst.hrst new file mode 100644 index 0000000..d0f113b --- /dev/null +++ b/Documentation/labs/subst.hrst @@ -0,0 +1,4 @@ +.. |LXR| replace:: LXR +.. _LXR: http://elixir.free-electrons.com/linux/latest/source + + diff --git a/Documentation/labs/vm.rst b/Documentation/labs/vm.rst new file mode 100644 index 0000000..8517308 --- /dev/null +++ b/Documentation/labs/vm.rst @@ -0,0 +1,157 @@ +===================== +Virtual Machine Setup +===================== + +Exercises are designed to run on a qemu based virtual machine. In +order to run the virtual machine you will need following packages: + +* build-essential +* qemu-system-x86 +* qemu-system-arm +* kvm +* python3 + +The virtual machine setup uses prebuild Yocto images that it downloads +from downloads.yocyoproject.org and a kernel image that it builds +itself. The following images are supported: + +* core-image-minimal-qemu +* core-image-minimal-dev-qemu +* core-image-sato-dev-qemu +* core-image-sato-qemu +* core-image-sato-sdk-qemu + +and can be selected from tools/labs/qemu/Makefile. + + +Starting the VM +--------------- + +The virtual machine scripts are available in tools/labs/qemeu and you +can can start the virtual machine by using the **boot** make target in +tools/labs: + +.. code-block:: shell + + ~/src/linux/tools/labs$ make boot + ARCH=x86 qemu/qemu.sh -kernel zImage.x86 -device virtio-serial \ + -chardev pty,id=virtiocon0 -device virtconsole,chardev=virtiocon0 \ + -net nic,model=virtio,vlan=0 -net tap,ifname=tap0,vlan=0,script=no,downscript=no\ + -drive file=rootfs.img,if=virtio,format=raw --append "root=/dev/vda console=hvc0" \ + --display none -s + char device redirected to /dev/pts/19 (label virtiocon0) + + +.. note:: To show the qemu console use "QEMU_DISPLAY=sdl make + boot". This will show the VGA output and will also give + access to the standard keyboard. + +Connecting to the VM +-------------------- + +Once the machine is booted you can connect to it on the serial port. A +link named *serial.pts* is created to the right emulated serial port +and you can use **minicom**, **picocom** to connect to the virtual +machine from the host: + +.. code-block:: shell + + $ minicom -D serial.pts + + Poky (Yocto Project Reference Distro) 2.3 qemux86 /dev/hvc0 + + qemux86 login: root + root@qemux86:~# + +Networking is also setup and you can use ssh to connect to the virtual +machine after finding out the allocated IP address: + +.. code-block:: shell + + $ mincom -D serial.pts + + Poky (Yocto Project Reference Distro) 2.3 qemux86 /dev/hvc0 + + qemux86 login: root + root@qemux86:~# ifconfig + eth0 Link encap:Ethernet HWaddr 52:54:00:12:34:56 + inet addr:172.20.0.6 Bcast:172.20.0.255 Mask:255.255.255.0 + UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 + RX packets:41 errors:0 dropped:0 overruns:0 frame:0 + TX packets:6 errors:0 dropped:0 overruns:0 carrier:0 + collisions:0 txqueuelen:1000 + RX bytes:7578 (7.4 KiB) TX bytes:1296 (1.2 KiB) + + lo Link encap:Local Loopback + inet addr:127.0.0.1 Mask:255.0.0.0 + inet6 addr: ::1%134535719/128 Scope:Host + UP LOOPBACK RUNNING MTU:65536 Metric:1 + RX packets:0 errors:0 dropped:0 overruns:0 frame:0 + TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 + collisions:0 txqueuelen:1000 + RX bytes:0 (0.0 B) TX bytes:0 (0.0 B) + + $ ssh root@172.20.0.6 + The authenticity of host '172.20.0.6 (172.20.0.6)' can't be established. + RSA key fingerprint is SHA256:CW1opJUHi4LDt1lnKjBVv12kXZ4s+8rreMBm5Jsdm00. + Are you sure you want to continue connecting (yes/no)? yes + Warning: Permanently added '172.20.0.6' (RSA) to the list of known hosts. + root@qemux86:~# + +.. attention:: The Yocto core-image-minimal-qemu does not include an + SSH server, so you will not able to connect via ssh if + you are using this image. + + +Connecting a debugger to the VM kernel +-------------------------------------- + +You can connect gdb to the running VM kernel and inspect the state of +the kernel by running the *gdb* target from tools/labs: + +.. code-block :: shell + + $ make gdb + ln -fs /home/tavi/src/linux/vmlinux vmlinux + gdb -ex "target remote localhost:1234" vmlinux + GNU gdb (Ubuntu 7.11.1-0ubuntu1~16.04) 7.11.1 + Copyright (C) 2016 Free Software Foundation, Inc. + License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html> + This is free software: you are free to change and redistribute it. + There is NO WARRANTY, to the extent permitted by law. Type "show copying" + and "show warranty" for details. + This GDB was configured as "x86_64-linux-gnu". + Type "show configuration" for configuration details. + For bug reporting instructions, please see: + <http://www.gnu.org/software/gdb/bugs/>. + Find the GDB manual and other documentation resources online at: + <http://www.gnu.org/software/gdb/documentation/>. + For help, type "help". + Type "apropos word" to search for commands related to "word"... + Reading symbols from vmlinux...done. + Remote debugging using localhost:1234 + 0xc13cf2f2 in native_safe_halt () at ./arch/x86/include/asm/irqflags.h:53 + 53asm volatile("sti; hlt": : :"memory"); + (gdb) bt + #0 0xc13cf2f2 in native_safe_halt () at ./arch/x86/include/asm/irqflags.h:53 + #1 arch_safe_halt () at ./arch/x86/include/asm/irqflags.h:95 + #2 default_idle () at arch/x86/kernel/process.c:341 + #3 0xc101f136 in arch_cpu_idle () at arch/x86/kernel/process.c:332 + #4 0xc106a6dd in cpuidle_idle_call () at kernel/sched/idle.c:156 + #5 do_idle () at kernel/sched/idle.c:245 + #6 0xc106a8c5 in cpu_startup_entry (state=<optimized out>) + at kernel/sched/idle.c:350 + #7 0xc13cb14a in rest_init () at init/main.c:415 + #8 0xc1507a7a in start_kernel () at init/main.c:679 + #9 0xc10001da in startup_32_smp () at arch/x86/kernel/head_32.S:368 + #10 0x00000000 in ?? () + (gdb) + +Rebuild the kernel image +------------------------ + +The kernel image is built the first time the VM is started. To rebuild +the kernel remove the **zImage** file and run the zImage target (or +start the VM again). + +.. add info about how to update the image diff --git a/README.rst b/README.rst new file mode 120000 index 0000000..a030cc1 --- /dev/null +++ b/README.rst @@ -0,0 +1 @@ +Documentation/labs/index.rst \ No newline at end of file diff --git a/tools/labs/.gitignore b/tools/labs/.gitignore new file mode 100644 index 0000000..a6940b6 --- /dev/null +++ b/tools/labs/.gitignore @@ -0,0 +1,7 @@ +skels +vmlinux +zImage +serial.pts +rootfs.img +core-image-minimal-*.ext4 +.modinst \ No newline at end of file diff --git a/tools/labs/Makefile b/tools/labs/Makefile new file mode 100644 index 0000000..d218a90 --- /dev/null +++ b/tools/labs/Makefile @@ -0,0 +1,42 @@ +KDIR=$(shell realpath $(PWD)/../..) + +LABS?=$(shell cd templates && find -mindepth 1 -maxdepth 1 -type d) +MODS=$(shell cd templates && find $(LABS) -mindepth 1 -name Kbuild | xargs dirname) +TODO?=0 + +include qemu/Makefile + +skels: + mkdir -p skels + cd templates && find $(LABS) -type f | xargs ./generate_skels.py --output ../skels --todo $(TODO) + rm -f skels/Kbuild + +skels/Kbuild: + echo "# autogenerated, do not edit " > $@ + echo "ccflags-y += -Wno-unused-function -Wno-unused-label -Wno-unused-variable " >> $@ + for i in $(shell cd skels && find -mindepth 1 -name Kbuild | xargs dirname); do echo "obj-m += $$i/" >> $@; done + +build: $(KCONFIG) skels/Kbuild + $(MAKE) -C $(KDIR) M=$(PWD)/skels ARCH=$(ARCH) modules + for i in $(shell find skels -name Makefile | xargs dirname); do $(MAKE) -C $$i; done + +TEMPDIR := $(shell mktemp -u) + +copy: $(YOCTO_IMAGE) + if [ -e qemu.mon ]; then exit 1; fi + mkdir $(TEMPDIR) + sudo mount -t ext4 -o loop $(YOCTO_IMAGE) $(TEMPDIR) + find skels -type f \( -name *.ko -or -executable \) | xargs sudo cp --parents -t $(TEMPDIR)/home/root || true + sudo umount $(TEMPDIR) + rmdir $(TEMPDIR) + +docs: + $(MAKE) -C $(KDIR) DOCBOOKS= SPHINXDIRS=labs htmldocs + +slides: + $(MAKE) -C $(KDIR) BUILDDIR=$(KDIR)/Documentation/output/slides DOCBOOKS= SPHINXDIRS=labs slides + +clean:: + rm -rf skels + +.PHONY: skels diff --git a/tools/labs/qemu/Makefile b/tools/labs/qemu/Makefile new file mode 100644 index 0000000..40e9181 --- /dev/null +++ b/tools/labs/qemu/Makefile @@ -0,0 +1,61 @@ +QEMU_DISPLAY ?= none +ARCH ?= x86 +ifeq ($(ARCH),x86) +b = b +endif +ZIMAGE=$(KDIR)/arch/$(ARCH)/boot/$(b)zImage +KCONFIG=$(KDIR)/.config + +YOCTO_URL=http://downloads.yoctoproject.org/releases/yocto/yocto-2.3/machines/qemu/qemu$(ARCH)/ +YOCTO_IMAGE=core-image-minimal-qemu$(ARCH).ext4 +#YOCTO_IMAGE=core-image-minimal-dev-qemu$(ARCH).ext4 +#YOCTO_IMAGE=core-image-sato-dev-qemu$(ARCH).ext4 +#YOCTO_IMAGE=core-image-sato-qemu$(ARCH).ext4 +#YOCTO_IMAGE=core-image-sato-sdk-qemu$(ARCH).ext4 + +QEMU_OPTS = -kernel $(ZIMAGE) \ + -device virtio-serial \ + -chardev pty,id=virtiocon0 -device virtconsole,chardev=virtiocon0 \ + -net nic,model=virtio,vlan=0 -net tap,ifname=tap0,vlan=0,script=no,downscript=no \ + -drive file=$(YOCTO_IMAGE),if=virtio,format=raw \ + --append "root=/dev/vda console=hvc0" \ + --display $(QEMU_DISPLAY) -s + +boot: .modinst tap0 + ARCH=$(ARCH) qemu/qemu.sh $(QEMU_OPTS) + +TEMPDIR := $(shell mktemp -u) + +$(KCONFIG): qemu/kernel_config.x86 + cp $^ $@ + $(MAKE) -C $(KDIR) oldnoconfig + +zImage: $(ZIMAGE) + +$(ZIMAGE): $(KCONFIG) + $(MAKE) -C $(KDIR) + $(MAKE) -C $(KDIR) modules + +.modinst: $(ZIMAGE) $(YOCTO_IMAGE) + mkdir $(TEMPDIR) + sudo mount -t ext4 -o loop $(YOCTO_IMAGE) $(TEMPDIR) + sudo $(MAKE) -C $(KDIR) modules_install INSTALL_MOD_PATH=$(TEMPDIR) + sudo umount $(TEMPDIR) + rmdir $(TEMPDIR) + sleep 1 && touch .modinst + +gdb: $(ZIMAGE) + gdb -ex "target remote localhost:1234" $(KDIR)/vmlinux + +$(YOCTO_IMAGE): + wget $(YOCTO_URL)/$(YOCTO_IMAGE) + sudo qemu/prepare-image.sh $(YOCTO_IMAGE) + +tap0: + qemu/create_net.sh $@ + +clean:: + rm -f .modinst + +.PHONY: clean tap0 + diff --git a/tools/labs/qemu/create_net.sh b/tools/labs/qemu/create_net.sh new file mode 100755 index 0000000..3f76547 --- /dev/null +++ b/tools/labs/qemu/create_net.sh @@ -0,0 +1,37 @@ +#!/bin/bash + +if test $# -ne 1; then + echo "Usage: $0 <device>" 1>&2 + echo " <device> must be tap0 or tap1" + exit 1 +fi + +device=$1 + +case "$device" in + "tap0") + subnet=172.213.0 + ;; + "tap1") + subnet=172.214.0 + ;; + *) + echo "Unknown device" 1>&2 + exit 1 + ;; +esac + +# If device doesn't exist add device. +if ! /sbin/ip link show dev "$device" > /dev/null 2>&1; then + sudo ip tuntap add mode tap user "$USER" dev "$device" +fi + +# Reconfigure just to be sure (even if device exists). +sudo /sbin/ip address flush dev "$device" +sudo /sbin/ip link set dev "$device" down +sudo /sbin/ip address add $subnet.1/24 dev "$device" +sudo /sbin/ip link set dev "$device" up + +mkdir -p $PWD/tftp + +sudo dnsmasq --enable-tftp --tftp-root=$PWD/tftp --no-resolv --no-hosts --bind-interfaces --interface $device -F $subnet.2,$subnet.20 -x dnsmasq.pid || true diff --git a/tools/labs/qemu/kernel_config.x86 b/tools/labs/qemu/kernel_config.x86 new file mode 100644 index 0000000..bb18e79 --- /dev/null +++ b/tools/labs/qemu/kernel_config.x86 @@ -0,0 +1,79 @@ +# CONFIG_64BIT is not set +# CONFIG_LOCALVERSION_AUTO is not set +# CONFIG_CROSS_MEMORY_ATTACH is not set +# CONFIG_USELIB is not set +CONFIG_BLK_DEV_INITRD=y +# CONFIG_COMPAT_BRK is not set +CONFIG_SLAB=y +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +CONFIG_SMP=y +# CONFIG_X86_EXTENDED_PLATFORM is not set +# CONFIG_SCHED_OMIT_FRAME_POINTER is not set +# CONFIG_X86_MCE is not set +# CONFIG_MICROCODE is not set +# CONFIG_COMPACTION is not set +# CONFIG_SECCOMP is not set +# CONFIG_RELOCATABLE is not set +# CONFIG_SUSPEND is not set +CONFIG_BINFMT_AOUT=y +CONFIG_BINFMT_MISC=y +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_UNIX=y +CONFIG_INET=y +# CONFIG_UEVENT_HELPER is not set +CONFIG_DEVTMPFS=y +# CONFIG_STANDALONE is not set +# CONFIG_PREVENT_FIRMWARE_BUILD is not set +CONFIG_VIRTIO_BLK=y +CONFIG_NETDEVICES=y +CONFIG_VIRTIO_NET=y +# CONFIG_DEVMEM is not set +CONFIG_VIRTIO_CONSOLE=y +# CONFIG_HW_RANDOM is not set +# CONFIG_HWMON is not set +# CONFIG_HID_A4TECH is not set +# CONFIG_HID_APPLE is not set +# CONFIG_HID_BELKIN is not set +# CONFIG_HID_CHERRY is not set +# CONFIG_HID_CHICONY is not set +# CONFIG_HID_CYPRESS is not set +# CONFIG_HID_EZKEY is not set +# CONFIG_HID_KENSINGTON is not set +# CONFIG_HID_LOGITECH is not set +# CONFIG_HID_MICROSOFT is not set +# CONFIG_HID_MONTEREY is not set +# CONFIG_USB_SUPPORT is not set +CONFIG_VIRTIO_PCI=y +# CONFIG_X86_PLATFORM_DEVICES is not set +# CONFIG_IOMMU_SUPPORT is not set +CONFIG_EXT4_FS=y +CONFIG_EXT4_FS_POSIX_ACL=y +CONFIG_EXT4_FS_SECURITY=y +CONFIG_EXT4_ENCRYPTION=y +# CONFIG_DNOTIFY is not set +# CONFIG_INOTIFY_USER is not set +CONFIG_PROC_KCORE=y +CONFIG_TMPFS=y +CONFIG_TMPFS_POSIX_ACL=y +CONFIG_CONFIGFS_FS=y +# CONFIG_MISC_FILESYSTEMS is not set +CONFIG_DEBUG_INFO=y +CONFIG_DEBUG_INFO_DWARF4=y +# CONFIG_ENABLE_WARN_DEPRECATED is not set +# CONFIG_ENABLE_MUST_CHECK is not set +# CONFIG_UNUSED_SYMBOLS is not set +# CONFIG_SECTION_MISMATCH_WARN_ONLY is not set +CONFIG_MAGIC_SYSRQ=y +CONFIG_DEBUG_KERNEL=y +CONFIG_DEBUG_SLAB=y +CONFIG_DEBUG_SLAB_LEAK=y +CONFIG_DEBUG_RT_MUTEXES=y +CONFIG_PROVE_LOCKING=y +CONFIG_DEBUG_ATOMIC_SLEEP=y +# CONFIG_FTRACE is not set +# CONFIG_X86_VERBOSE_BOOTUP is not set +# CONFIG_X86_DEBUG_FPU is not set +CONFIG_CRYPTO_ECHAINIV=y +# CONFIG_VIRTUALIZATION is not set diff --git a/tools/labs/qemu/prepare-image.sh b/tools/labs/qemu/prepare-image.sh new file mode 100755 index 0000000..459fa74 --- /dev/null +++ b/tools/labs/qemu/prepare-image.sh @@ -0,0 +1,28 @@ +#!/bin/bash + +size=$(stat -c%s $1) +if [ $size -lt 50000000 ]; then + e2fsck -f $1 + resize2fs $1 50M +fi + +TMP=$(mktemp -d) + +mount -t ext4 -o loop $1 $TMP + +# add console +echo "hvc0:12345:respawn:/sbin/getty 115200 hvc0" >> $TMP/etc/inittab + +# add more vty +cat >> $TMP/etc/inittab <<EOF +2:12345:respawn:/sbin/getty 38400 tty2 +3:12345:respawn:/sbin/getty 38400 tty3 +4:12345:respawn:/sbin/getty 38400 tty4 +5:12345:respawn:/sbin/getty 38400 tty5 +EOF + +# enable networking +echo -e "auto eth0\niface eth0 inet dhcp" >> $TMP/etc/network/interfaces + +sudo umount $TMP +rmdir $TMP diff --git a/tools/labs/qemu/qemu.sh b/tools/labs/qemu/qemu.sh new file mode 100755 index 0000000..329ef34 --- /dev/null +++ b/tools/labs/qemu/qemu.sh @@ -0,0 +1,21 @@ +#!/bin/bash +# +# This script runs qemu and creates a symbolic link named serial.pts +# to the qemu serial console (pts based). Because the qemu pts +# allocation is dynamic, it is preferable to have a stable path to +# avoid visual inspection of the qemu output when connecting to the +# serial console. + +case $ARCH in + x86) + qemu=qemu-system-i386 + ;; + arm) + qemu=qemu-system-arm + ;; +esac + +echo info chardev | nc -U -l qemu.mon | egrep -o "/dev/pts/[0-9]*" | xargs -I PTS ln -fs PTS serial.pts & +$qemu "$@" -monitor unix:qemu.mon +rm qemu.mon +rm serial.pts diff --git a/tools/labs/templates/generate_skels.py b/tools/labs/templates/generate_skels.py new file mode 100755 index 0000000..17a6ea6 --- /dev/null +++ b/tools/labs/templates/generate_skels.py @@ -0,0 +1,55 @@ +#!/usr/bin/python3 -u + +import argparse, fnmatch, glob, os.path, re, sys, shutil + +parser = argparse.ArgumentParser(description='Generate skeletons sources from full sources') +parser.add_argument('paths', metavar='path', nargs='+', help='list of files to process') +parser.add_argument('--output', help='output dir to copy processed files') +parser.add_argument('--todo', type=int, help='don\'t remove TODOs less then this', default=1) +args = parser.parse_args() + +def process_file(p, pattern): + f = open(p, "r") + g = open(os.path.join(args.output, p), "w") + skip_lines = 0 + for l in f.readlines(): + if skip_lines > 0: + skip_lines -= 1 + m = re.search(pattern, l) + if m : + l = "%s%s%s\n" % (m.group(1), m.group(2), m.group(4)) + g.write(l) + continue + m = re.search(pattern, l) + if m: + todo=1 + if m.group(2): + todo = int(m.group(2)) + if todo >= args.todo: + if m.group(3): + skip_lines = int(m.group(3)) + else: + skip_lines = 1 + l = "%s%s%s\n" % (m.group(1), m.group(2), m.group(4)) + g.write(l) + +for p in args.paths: + print("skel %s" % (p), sep = '') + name=os.path.basename(p) + try: + os.makedirs(os.path.join(args.output, os.path.dirname(p))) + except: + pass + + copy = False + if name == "Kbuild" or name == "Makefile": + pattern="(^#\s*TODO)([0-9]*)\/?([0-9]*)(:.*)" + elif fnmatch.fnmatch(name, '*.c') or fnmatch.fnmatch(name, '*.h'): + pattern="(.*/\*\s*TODO)([ 0-9]*)/?([0-9]*)(:.*)" + else: + copy = True + + if copy: + shutil.copyfile(p, os.path.join(args.output, p)) + else: + process_file(p, pattern) -- 2.7.4 -- To unsubscribe from this list: send the line "unsubscribe linux-doc" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
