docs/ci/bare-metal.rst

   1 Bare-metal CI
   2 =============
   3
   4 The bare-metal scripts run on a system with gitlab-runner and docker,
   5 connected to potentially multiple bare-metal boards that run tests of
   6 Mesa.  Currently only "fastboot" and "ChromeOS Servo" devices are
   7 supported.
   8
   9 In comparison with LAVA, this doesn't involve maintaining a separate
  10 webservice with its own job scheduler and replicating jobs between the
  11 two.  It also places more of the board support in git, instead of
  12 webservice configuration.  On the other hand, the serial interactions
  13 and bootloader support are more primitive.
  14
  15 Requirements (fastboot)
  16 -----------------------
  17
  18 This testing requires power control of the DUTs by the gitlab-runner
  19 machine, since this is what we use to reset the system and get back to
  20 a pristine state at the start of testing.
  21
  22 We require access to the console output from the gitlab-runner system,
  23 since that is how we get the final results back from the tests.  You
  24 should probably have the console on a serial connection, so that you
  25 can see bootloader progress.
  26
  27 The boards need to be able to have a kernel/initramfs supplied by the
  28 gitlab-runner system, since the initramfs is what contains the Mesa
  29 testing payload.
  30
  31 The boards should have networking, so that (in a future iteration of
  32 this code) we can extract the dEQP .xml results to artifacts on
  33 gitlab.
  34
  35 Requirements (servo)
  36 --------------------
  37
  38 For servo-connected boards, we can use the EC connection for power
  39 control to reboot the board.  However, loading a kernel is not as easy
  40 as fastboot, so we assume your bootloader can do TFTP, and that your
  41 gitlab-runner mounts the runner's tftp directory specific to the board
  42 at /tftp in the container.
  43
  44 Since we're going the TFTP route, we also use NFS root.  This avoids
  45 packing the rootfs and sending it to the board as a ramdisk, which
  46 means we can support larger rootfses (for piglit or tracie testing),
  47 at the cost of needing more storage on the runner.
  48
  49 Telling the board about where its TFTP and NFS should come from is
  50 done using dnsmasq on the runner host.  For example, this snippet in
  51 the dnsmasq.conf.d in the google farm, with the gitlab-runner host we
  52 call "servo"::
  53
  54   dhcp-host=1c:69:7a:0d:a3:d3,10.42.0.10,set:servo
  55
  56   # Fixed dhcp addresses for my sanity, and setting a tag for
  57   # specializing other DHCP options
  58   dhcp-host=a0:ce:c8:c8:d9:5d,10.42.0.11,set:cheza1
  59   dhcp-host=a0:ce:c8:c8:d8:81,10.42.0.12,set:cheza2
  60
  61   # Specify the next server, watch out for the double ',,'.  The
  62   # filename didn't seem to get picked up by the bootloader, so we use
  63   # tftp-unique-root and mount directories like
  64   # /srv/tftp/10.42.0.11/jwerner/cheza as /tftp in the job containers.
  65   tftp-unique-root
  66   dhcp-boot=tag:cheza1,cheza1/vmlinuz,,10.42.0.10
  67   dhcp-boot=tag:cheza2,cheza2/vmlinuz,,10.42.0.10
  68
  69   dhcp-option=tag:cheza1,option:root-path,/srv/nfs/cheza1
  70   dhcp-option=tag:cheza2,option:root-path,/srv/nfs/cheza2
  71
  72 Setup
  73 -----
  74
  75 Each board will be registered in fd.o gitlab.  You'll want something
  76 like this to register a fastboot board:
  77
  78 .. code-block:: console
  79
  80   sudo gitlab-runner register \
  81        --url https://gitlab.freedesktop.org \
  82        --registration-token $1 \
  83        --name MY_BOARD_NAME \
  84        --tag-list MY_BOARD_TAG \
  85        --executor docker \
  86        --docker-image "alpine:latest" \
  87        --docker-volumes "/dev:/dev" \
  88        --docker-network-mode "host" \
  89        --docker-privileged \
  90        --non-interactive
  91
  92 For a servo board, you'll need to also volume mount the board's NFS
  93 root dir at /nfs and TFTP kernel directory at /tftp.
  94
  95 The registration token has to come from a fd.o gitlab admin going to
  96 https://gitlab.freedesktop.org/admin/runners
  97
  98 The name scheme for Google's lab is google-freedreno-boardname-n, and
  99 our tag is something like google-freedreno-db410c.  The tag is what
 100 identifies a board type so that board-specific jobs can be dispatched
 101 into that pool.
 102
 103 We need privileged mode and the /dev bind mount in order to get at the
 104 serial console and fastboot USB devices (--device arguments don't
 105 apply to devices that show up after container start, which is the case
 106 with fastboot, and the servo serial devices are acctually links to
 107 /dev/pts).  We use host network mode so that we can (in the future)
 108 spin up a server to collect XML results for fastboot.
 109
 110 Once you've added your boards, you're going to need to add a little
 111 more customization in ``/etc/gitlab-runner/config.toml``.  First, add
 112 ``concurrent = <number of boards>`` at the top ("we should have up to
 113 this many jobs running managed by this gitlab-runner").  Then for each
 114 board's runner, set ``limit = 1`` ("only 1 job served by this board at a
 115 time").  Finally, add the board-specific environment variables
 116 required by your bare-metal script, something like::
 117
 118   [[runners]]
 119     name = "google-freedreno-db410c-1"
 120     environment = ["BM_SERIAL=/dev/ttyDB410c8", "BM_POWERUP=google-power-up.sh 8", "BM_FASTBOOT_SERIAL=15e9e390"]
 121
 122 Once you've updated your runners' configs, restart with ``sudo service
 123 gitlab-runner restart``