[Macchiato] EDKII grub boot fails with PCIe init
Frederik Lotter
frederik.lotter at netronome.com
Wed Mar 28 11:00:03 BST 2018
Hi Marcin,
On Fri, Mar 23, 2018 at 1:38 PM, Frederik Lotter <
frederik.lotter at netronome.com> wrote:
>
>
> On Fri, Mar 23, 2018 at 1:02 PM, Marcin Wojtas <mw at semihalf.com> wrote:
>
>> Frederik,
>>
>>
>> 2018-03-23 11:02 GMT+01:00 Marcin Wojtas <mw at semihalf.com>:
>> > Hi Frederik,
>> >
>> > 2018-03-23 8:07 GMT+01:00 Frederik Lotter <
>> frederik.lotter at netronome.com>:
>> >>
>> >>
>> >>
>> >> On Fri, Mar 16, 2018 at 5:16 PM, Marcin Wojtas <mw at semihalf.com>
>> wrote:
>> >>>
>> >>> Hi Frederik,
>> >>>
>> >>> 2018-03-16 14:04 GMT+01:00 Frederik Lotter <
>> frederik.lotter at netronome.com>:
>> >>> >
>> >>> > On Wed, Mar 14, 2018 at 10:56 AM, Frederik Lotter <
>> frederik.lotter at netronome.com> wrote:
>> >>> >>
>> >>> >> On Wed, Mar 14, 2018 at 10:13 AM, Frederik Lotter <
>> frederik.lotter at netronome.com> wrote:
>> >>> >>>
>> >>> >>> On Tue, Mar 13, 2018 at 6:23 PM, Marcin Wojtas <mw at semihalf.com>
>> wrote:
>> >>> >>>>
>> >>> >>>> Frederic,
>> >>> >>>>
>> >>> >>>> >>> Please use the size I suggested for the 'reg' property
>> >>> >>>> >>
>> >>> >>>> >>
>> >>> >>>> >> Sorry I missed that:
>> >>> >>>> >>
>> >>> >>>> >> [ 1.463413] PCI: OF: host bridge /cp0/pcie at 0xe0000000
>> ranges:
>> >>> >>>> >> [ 1.463427] PCI: OF: IO 0xeff00000..0xeff0ffff ->
>> 0x00000000
>> >>> >>>> >> [ 1.463435] PCI: OF: MEM 0xc0000000..0xdfffffff ->
>> 0xc0000000
>> >>> >>>> >> [ 1.463442] PCI: OF: MEM 0x800000000..0x8ffffffff ->
>> 0x800000000
>> >>> >>>> >> [ 1.463481] pci-host-generic e0000000.pcie: ECAM at [mem
>> >>> >>>> >> 0xe0000000-0xefefffff] for [bus 00-fe]
>> >>> >>>> >> [ 1.463525] pci-host-generic e0000000.pcie: PCI host
>> bridge to bus
>> >>> >>>> >> 0000:00
>> >>> >>>> >> [ 1.463531] pci_bus 0000:00: root bus resource [bus 00-fe]
>> >>> >>>> >> [ 1.463536] pci_bus 0000:00: root bus resource [io
>> 0x0000-0xffff]
>> >>> >>>> >> [ 1.463541] pci_bus 0000:00: root bus resource [mem
>> >>> >>>> >> 0xc0000000-0xdfffffff]
>> >>> >>>> >> [ 1.463547] pci_bus 0000:00: root bus resource [mem
>> >>> >>>> >> 0x800000000-0x8ffffffff]
>> >>> >>>> >>
>> >>> >>>> >>
>> >>> >>>> >> So I assume this works, and I am super grateful. I will test
>> it tomorrow
>> >>> >>>> >> with our Smart NIC.
>> >>> >>>> >>
>> >>> >>>>
>> >>> >>>> Please also check pure branch (unmodified DT) as requested in my
>> >>> >>>> previous email . According to the bootlog the stall was observed
>> after
>> >>> >>>> pcie driver init and without latest patch I mentioned. So I'd
>> like to
>> >>> >>>> make sure that we're on the same side here.
>> >>> >>>
>> >>> >>>
>> >>> >>> I attached two logs files. The only difference is the one also
>> uses an initrd.
>> >>> >>>
>> >>> >>> I now have a pcie card in - and it looks much better.
>> >>> >>>
>> >>> >>> I am starting to feel like this whole exercise is going to
>> discover something unrelated and stupid I have done, so I apologize in
>> advance. Hopefully the journey will help someone else as well.
>> >>> >>>
>> >>> >>> Any idea what could be failing? I seem to be some issue mounting
>> the root filesystem (or at the same time). In the initrd case it seems to
>> complain about mmcblk0 (the onboard flash which I do not use).
>> >>> >>>
>> >>> >>> I attach my grub file just for in case - i am new to grub (Note I
>> hacked 4 entries towards the end).
>> >>> >>
>> >>> >>
>> >>> >> Here are 3 more logs, I will not send more - but I thought perhaps
>> this could complete the picture.
>> >>> >>
>> >>> >> 1. One without a PCIe card - successful PCIe init
>> >>> >>
>> >>> >> 2. One with a card, but with the PCIe probe causing the stall
>> (this is not often seen)
>> >>> >>
>> >>> >> 3. The other stall, after successful PCIe init
>> >>> >
>> >>> >
>> >>> > Hi Marcin,
>> >>> >
>> >>> > I am sure you are quite busy, and I really appreciate all the help
>> I got from you.
>> >>> >
>> >>> > Please will you have a look at the last two emails (and the
>> attachments) I've sent you, once you have time again. If there is anything
>> I can do that will help you, just let me know. I really need to get this
>> working reliably for us to proceed with the EFI boot route, and since we
>> really need to support generic netboot/ISO installs and images for CentOS
>> and Ubuntu, I think this must be the best way forward.
>> >>> >
>> >>>
>> >>> I took a look at your logs and it all looks a bit strange. Is it pure
>> >>> v4.16-rc5? If yes, can you avoid grub and boot directly from shell?
>> >>
>> >>
>> >> Could you give me a hint how I do this? I am very new to EDK/grub?
>> >
>> > When booting, hit escape. Go to "Boot Manager" and then to "UEFI
>> > Shell". Navigate to partition, which comprises your Linux Image (let's
>> > assume it's FS0):
>> > Shell> fs0:
>> > FS0:\> Image <commandline arguments>
>> >
>> > Example:
>> >
>> > https://pastebin.com/rzk9uzAx
>>
>
This does not work for me because EDK does not have a working alias for
mmcblk1p3. In the past I had the EFI System partition last, but now I have
the Linux Filesystem last (not sure if this is confusing EDK)
Mapping table
FS0: Alias(s):HD4c:;BLK5:
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,000078F20000000000)/SD(0x0)
/HD(2,GPT,82F4BCFA-CC74-4966-A5BB-3206A1D4B694,0x11000,0x64000)
FS1: Alias(s):F0:;BLK7:
VenMsg(06ED4DD0-FF78-11D3-BDC4-00A0C94053D1,0200000000000000)
BLK0: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,00006EF00000000000)/eMMC(0x
0)/Ctrl(0x0)
BLK1: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,00006EF00000000000)/eMMC(0x
0)/Ctrl(0x1)
BLK2: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,00006EF00000000000)/eMMC(0x
0)/Ctrl(0x2)
BLK3: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,000078F20000000000)/SD(0x0)
BLK4: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,000078F20000000000)/SD(0x0)
/HD(1,GPT,3AA5DACC-F435-43FF-828F-A6374A093D21,0x1000,0x10000)
BLK6: Alias(s):
VenHw(0D51905B-B77E-452A-A2C0-ECA0CC8D514A,000078F20000000000)/SD(0x0)
/HD(3,GPT,7E0D3DE0-AD08-4599-9AA3-164B3D83EA0D,0x75000,0x6EDBDF)
The Image is on BLK6: which I cannot change to - nothing happens if I type
"BLK6:".
>
>> >
>> >
>> >>
>> >>>
>> >>>
>> >>> Can you share your grub? I'd like to test it in my setup.
>>
>
I think the easiest will be if you flash this image on your sdcard.
https://drive.google.com/open?id=1qqGfgFw-rcc1d09a2F-BoRgmBnmmqrw5
So I have made the whole image from scratch:
Userspace: 18.04 Ubuntu-server squashfs
Kernel: mainline 4.16 rc7 (arm64 generic deconfig)
mmcblk1p1: uboot/EDK image
mmcblk1p2: ESP
mmcblk1p3: Ubuntu userspace
The way I did it was to get uboot based boot up, and then I do:
apt-get install grub-efi-arm64
update-grub (creates the grub.cfg)
grub-install /dev/mmcblk1 --removable (without removable I get some
efbootmgr errors related to efivars)
Either I am the only one trying EFI with custom images, or I am making an
obvious mistake (which is likely). However, I have done it so many times
now with different permutations, that I have exhausted all possible
combinations.
I would really appreciate it so much if you could just boot the image and
inspect it locally.
(The flash-image.bin was build again from vanilla sources including the
latest commits on the usual branches).
>>
>> >>
>> >> Please find the content of my EFI partition and /boot/grub attached. I
>> will really appreciate if you could give it a go.
>> >>
>> >> I got this grub through apt-get for Ubuntu.
>> >
>> > Thanks, will try it.
>> >
>>
>> Where do you take your initrd images from? Any chance you expose your
>> /boot directory as a tarball, I'd like to run your images as well.
>>
>
> I really appreciate your effort in helping.
>
> Sadly I tried to update and start fresh, so I don't currently have a
> working environment I can share, because I cannot get the vanilla mainline
> kernel to work on the board (I posted another question about this).
>
> I will send you my entire setup ASAP.
>
>
>> Thanks,
>> Marcin
>>
>> >>
>> >> apt-get install grub-efi-arm64
>> >>
>> >> Info below:
>> >>
>> >> root at localhost:/boot# apt-get install grub-efi-arm64
>> >> Reading package lists... Done
>> >> Building dependency tree
>> >> Reading state information... Done
>> >> grub-efi-arm64 is already the newest version (2.02~beta2-36ubuntu3.17).
>> >> 0 upgraded, 0 newly installed, 0 to remove and 48 not upgraded.
>> >> root at localhost:/boot# uname -a
>> >> Linux localhost.localdomain 4.16.0-rc5-mbcin-netronome-2-dirty #2 SMP
>> PREEMPT Mon Mar 12 14:40:25 UTC 2018 aarch64 aarch64 aarch64 GNU/Linux
>> >> root at localhost:/boot# cat /etc/os-release
>> >> NAME="Ubuntu"
>> >> VERSION="16.04.3 LTS (Xenial Xerus)"
>> >> ID=ubuntu
>> >> ID_LIKE=debian
>> >> PRETTY_NAME="Ubuntu 16.04.3 LTS"
>> >> VERSION_ID="16.04"
>> >> HOME_URL="http://www.ubuntu.com/"
>> >> SUPPORT_URL="http://help.ubuntu.com/"
>> >> BUG_REPORT_URL="http://bugs.launchpad.net/ubuntu/"
>> >> VERSION_CODENAME=xenial
>> >> UBUNTU_CODENAME=xenial
>> >>
>> >>>
>> >>>
>> >>> Can you please download:
>> >>> https://d-i.debian.org/daily-images/arm64/20180314-02:10/net
>> boot/mini.iso
>> >>> burn on stick and install? It's very easy and I use it in my GPU setup
>> >>
>> >>
>> >> OK I will try this but first I want to redo my environment from
>> scratch so make sure I have not messed up anything.
>> >>
>> >>>
>> >>> on MacchiatoBin.
>> >>>
>> >>> Thanks,
>> >>> Marcin
>> >>>
>> >>> >
>> >>> >>
>> >>> >>>
>> >>> >>>>
>> >>> >>>> >
>> >>> >>>> > This doesn't tell you all that much, to be honest. But at
>> least the
>> >>> >>>> > numbers look sane now, and appear to match the UEFI
>> configuration.
>> >>> >>>> >
>> >>> >>>> >> However, we are building a product that obviously requires
>> long term
>> >>> >>>> >> maintenance, so may I please get your input on a strategy
>> with this?
>> >>> >>>> >>
>> >>> >>>> >> If we decide to stick with this driver, would it be easy for
>> things to
>> >>> >>>> >> become disjointed?
>> >>> >>>> >>
>> >>> >>>> >> The hope with going the EFI route is that we could boot
>> "generic" Ubuntu and
>> >>> >>>> >> CentOS installs, so I guess as long as we keep the DT and the
>> EFKII snapshot
>> >>> >>>> >> in sync on our side, the risk is low.
>> >>> >>>> >>
>> >>> >>>> >
>> >>> >>>> > I'm afraid you are getting caught in the middle of a
>> philosophical
>> >>> >>>> > debate here: many engineers that are involved with the Marvell
>> support
>> >>> >>>> > in Linux feel that a device tree is not something that should
>> be
>> >>> >>>> > supported long term, and needs to be bundled with the OS. Over
>> the
>> >>> >>>> > last couple of kernel releases, the Marvell 8040 support was
>> changed
>> >>> >>>> > in a non-backward compatible manner numerous times.
>> >>> >>>> >
>> >>> >>>>
>> >>> >>>> I think current DT should work with everything >= v4.12. So far
>> >>> >>>> multiple users were able to install debian with recent fixes, I
>> >>> >>>> suggest first making sure, what possibly can happen that your
>> setup
>> >>> >>>> behaves differently. Switching to a8k-ecam-pcie driver is a nice
>> idea,
>> >>> >>>> but I'm not sure the distros using DT have it.
>> >>> >>>>
>> >>> >>>> > This conflicts badly with the idea that the firmware provides
>> the
>> >>> >>>> > hardware description (using DT or ACPI), and that the contract
>> with
>> >>> >>>> > the OS is kept by both sides for longer than a single release.
>> >>> >>>> >
>> >>> >>>> > So I cannot really answer that question, unfortunately. If you
>> don't
>> >>> >>>> > intend to use the onboard network controller, you could go the
>> ACPI
>> >>> >>>> > route, I guess.
>> >>> >>>>
>> >>> >>>> FYI. on-board network ACPI support is being upstreamed to the
>> Centos.
>> >>> >>>>
>> >>> >>>> >
>> >>> >>>> > Another problem is that none of this UEFI/ACPI support is
>> upstream in
>> >>> >>>> > the Tianocore project, and trying random trees left and right
>> doesn't
>> >>> >>>> > really help when assessing whether a platform is suitable as a
>> long
>> >>> >>>> > term investment.
>> >>> >>>> >
>> >>> >>>>
>> >>> >>>> There's only single branch recommended in the MacchiatoBin wiki,
>> I
>> >>> >>>> wouldn't call it 'random'. Entire branch is supposed to land
>> >>> >>>> eventually in the Tianocore and become the only support. Before
>> end of
>> >>> >>>> year ~50 patches got there, still some bits are missing, but I
>> think
>> >>> >>>> we're not that far from desired point. I really want to push it
>> but
>> >>> >>>> still it requires time I'm personally short of, so I'll
>> appreciate
>> >>> >>>> understanding.
>> >>> >>>>
>> >>> >>>> Thanks,
>> >>> >>>> Marcin
>> >>> >>>>
>> >>> >>>> >
>> >>> >>>> >
>> >>> >>>> >> For example, using the same DT with uboot, it fails:
>> >>> >>>> >>
>> >>> >>>> >> [ 0.294942] sysfs: cannot create duplicate filename
>> >>> >>>> >> '/bus/platform/devices/e0000000.pcie'
>> >>> >>>> >> [ 0.294950] CPU: 2 PID: 1 Comm: swapper/0 Not tainted
>> >>> >>>> >> 4.16.0-rc5-mbcin-netronome-2-dirty #2
>> >>> >>>> >> [ 0.294952] Hardware name: Marvell 8040 MACHIATOBin (DT)
>> >>> >>>> >> [ 0.294955] Call trace:
>> >>> >>>> >> [ 0.294967] dump_backtrace+0x0/0x150
>> >>> >>>> >> [ 0.294970] show_stack+0x14/0x20
>> >>> >>>> >> [ 0.294976] dump_stack+0x98/0xbc
>> >>> >>>> >> [ 0.294980] sysfs_warn_dup+0x60/0x78
>> >>> >>>> >> [ 0.294983] sysfs_do_create_link_sd.isra.0+0xd8/0xe0
>> >>> >>>> >> [ 0.294986] sysfs_create_link+0x20/0x40
>> >>> >>>> >> [ 0.294990] bus_add_device+0x88/0x148
>> >>> >>>> >> [ 0.294993] device_add+0x394/0x568
>> >>> >>>> >> [ 0.294997] of_device_add+0x5c/0x70
>> >>> >>>> >> [ 0.295000] of_platform_device_create_pdata+0x80/0xd0
>> >>> >>>> >> [ 0.295003] of_platform_bus_create+0xdc/0x300
>> >>> >>>> >> [ 0.295006] of_platform_bus_create+0x11c/0x300
>> >>> >>>> >> [ 0.295008] of_platform_populate+0x4c/0xb0
>> >>> >>>> >> [ 0.295014] of_platform_default_populate_init+0xa4/0xc0
>> >>> >>>> >> [ 0.295017] do_one_initcall+0x38/0x120
>> >>> >>>> >> [ 0.295020] kernel_init_freeable+0x134/0x1d4
>> >>> >>>> >> [ 0.295025] kernel_init+0x10/0x100
>> >>> >>>> >> [ 0.295028] ret_from_fork+0x10/0x18
>> >>> >>>> >>
>> >>> >>>> >> So I think this confirms that the pcie setup is different
>> between EDKII and
>> >>> >>>> >> uboot (unless I am doing something stupid here).
>> >>> >>>> >>
>> >>> >>>> >
>> >>> >>>> > It looks like you have two copies of the pcie node here, no?
>> >>> >>>
>> >>> >>>
>> >>> >>
>> >>> >
>> >>
>> >>
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.einval.com/pipermail/macchiato/attachments/20180328/3f70573d/attachment-0001.html>
More information about the Macchiato
mailing list