Nishanth Menon [Sat, 4 Nov 2023 07:21:49 +0000 (02:21 -0500)]
configs: j7200_evm_a72_defconfig: Switch to bootstd
Switch to using bootstd. Note with this change, we will stop using
distro_bootcmd and instead depend entirely on bootflow method of
starting the system up.
Signed-off-by: Nishanth Menon <nm@ti.com> Reviewed-by: Neha Malcom Francis <n-francis@ti.com>
Nishanth Menon [Sat, 4 Nov 2023 07:21:48 +0000 (02:21 -0500)]
configs: j7200: Remove HBMC_AM654 config
Kernel commit 1b77265626a4 ("arm64: dts: ti: k3-j7200-mcu-wakeup: Add
HyperBus node") was merged to kernel without its dependent patch [1].
Similar fix is needed in U-Boot, and hbmc currently breaks boot. Till
this gets fixed in U-Boot, disable the config by default so that the
hbmc probe that happens in board/ti/j721e/evm.c will not take place
and lead to boot failure.
This is similar to the approach in commit 5b2671594b80 ("configs:
j721e: Remove HBMC_AM654 config"), introduced to j7200 evm platform.
Nishanth Menon [Sat, 4 Nov 2023 07:21:47 +0000 (02:21 -0500)]
arm: mach-k3: j721e: Improve support for UDA FS
Commit 5019170970ad ("arch: arm: mach-k3: j721e: add support for UDA
FS") introduced basic UDA FS support, however, we can Take approach
similar to commit 0f1c1e8b368b ("arm: mach-k3: am625: Add support for
UDA FS"). While boot partition support with EMMC boot is useful, it is
constrained by the size of boot hardware partition itself.
In the case of K3 devices, tispl images can contain OP-TEE images that
can substantially vary in size and the u-boot image itself can vary over
time as we enable various features.
So use the CSD information in the case of EMMC_BOOT configuration being
enabled to pick boot partition or UDA FS mode operation to pick.
If EMMC_BOOT is disabled, then depend on filesystem configuration to
pick data from UDA.
Nishanth Menon [Sat, 4 Nov 2023 07:21:44 +0000 (02:21 -0500)]
arm: mach-k3: Kconfig: Introduce a symbol to indicate J7200
J7200 shares quite a few characteristics with J721E. However a few sets
are different. Introduce a Kconfig to differentiate the two to allow for
new boards to be introduced in a seamless manner.
Nishanth Menon [Sat, 4 Nov 2023 07:21:43 +0000 (02:21 -0500)]
configs: j721e_evm_a72_defconfig: Switch to bootstd
Switch to using bootstd. Note with this change, we will stop using
distro_bootcmd and instead depend entirely on bootflow method of
starting the system up.
Signed-off-by: Nishanth Menon <nm@ti.com> Reviewed-by: Neha Malcom Francis <n-francis@ti.com>
Nishanth Menon [Sat, 4 Nov 2023 02:45:11 +0000 (21:45 -0500)]
arm: mach-k3: Move TI dummy keys out of board folder
This file is used to emulate customer keys on TI development board
ecosystems, move it out of board/ directory and into mach-k3. And
change the relative paths to absolute paths in the binman paths.
While at it, drop the reference in verdin-binman file which is
redundant.
Nishanth Menon [Sat, 4 Nov 2023 02:45:10 +0000 (21:45 -0500)]
arm: mach-k3: Move K3 degenerate keys out of board folder
This file is common for all of K3, move it out of board/ directory and
into mach-k3. And change the relative paths to absolute paths in the
binman paths.
While at it, drop the reference in verdin-binman file which is
redundant.
Signed-off-by: Nishanth Menon <nm@ti.com> Reviewed-by: Francesco Dolcini <francesco.dolcini@toradex.com> Reviewed-by: Andrew Davis <afd@ti.com>
Andrew Davis [Tue, 14 Nov 2023 15:59:49 +0000 (09:59 -0600)]
arm: mach-k3: Remove incorrect checks for SPL build
The kconfig option SPL means this build supports SPL but not that
this build is SPL, nor that this build is the SPL running on R5.
For options that are for R5 SPL use CPU_V7R.
Andrew Davis [Wed, 1 Nov 2023 20:35:30 +0000 (15:35 -0500)]
arm: mach-k3: j721s2: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:29 +0000 (15:35 -0500)]
arm: mach-k3: am62ax: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:28 +0000 (15:35 -0500)]
arm: mach-k3: am62x: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:27 +0000 (15:35 -0500)]
arm: mach-k3: am64x: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:26 +0000 (15:35 -0500)]
arm: mach-k3: am65x: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:25 +0000 (15:35 -0500)]
arm: mach-k3: j721e: Move board selection to mach-k3
Currently each set of board targets from a vendor is selected inside
the board directory for that vendor. This has the problem of multiple
targets, one from each vendor, being selectable at the same time.
For instance you can select both TARGET_AM654_A53_EVM and
TARGET_IOT2050_A53 in the same build.
To fix this we need to move the target board choice to a common location
for each parent SoC selection. Do this in arch/arm/mach-k3.
Andrew Davis [Wed, 1 Nov 2023 20:35:24 +0000 (15:35 -0500)]
board: ti: Add dependency from TARGET selection to SOC
Currently the K3 selection for TARGET boards does not depend on the SoC
for which it is based. This leds to the odd ability to select for instance
both SOC_K3_AM625 and TARGET_J721E_A72_EVM.
To fix this the target choice should depend on the matching SOC config.
Signed-off-by: Andrew Davis <afd@ti.com> Reviewed-by: Neha Malcom Francis <n-francis@ti.com>
scsi: set dma direction to NONE for TEST UNIT READY
SCSI device scan code was executing TEST UNIT READY command without
explicitly setting dma direction in struct scsi_cmd to NONE, so command
was passed to driver with dma direction set to DMA_FROM_DEVICE,
inherited from older usage.
With WDC SDINDDH6-64G ufs device, that caused TEST UNIT READY to
return error.
Fix that, by explicitly setting dma direction to NONE for
TEST UNIT READY, and restoring it back DMA_FROM_DEVICE for the
following READ CAPACITY.
Signed-off-by: Nikita Yushchenko <nikita.yoush@cogentembedded.com> Reviewed-by: Marek Vasut <marex@denx.de>
Tom Rini [Sat, 18 Nov 2023 20:52:53 +0000 (15:52 -0500)]
Merge tag 'efi-next-18112023' of https://source.denx.de/u-boot/custodians/u-boot-tpm into next
EFI HTTP Boot is currently supported by using a combination of
wget, blkmap and bootefi commands. The user has to download the
image, mount it using blkmap and then execute the efi installer
using bootefi.
This series simplifies the user experience. Instead of doing all the
steps manually, users can now enable a new Kconfig (EFI_HTTP_BOOT)
which will select wget, blkmap and dns options. They can then use
efidebug command to add a boot option for the EFI Bootmanager using
=> efidebug boot add -u 3 netinst http://<path>
=> efidebug boot order 3
=> bootefi bootmgr
The boot manager will automatically download and mount the image.
Once it's mounted it will locate and launch the installer.
It's worth noting that this rarely fails, but the reason is irrelevant
to the current patchset. More information can be found here
https://lore.kernel.org/u-boot/CAOMZO5CoduEgwgdQiybmoKh6qQZOezUtRRQO4ecaGdZBBz5dDw@mail.gmail.c
om/
The tl;dr is that wget sometimes fails to download the file correctly
or set the size env variables. We expect all these to be solved once
LWIP is stable and pulled
Masahisa Kojima [Fri, 10 Nov 2023 04:25:41 +0000 (13:25 +0900)]
cmd: efidebug: add uri device path
This adds the URI device path option for 'boot add' subcommand.
User can add the URI load option for downloading ISO image file
or EFI application through network. Currently HTTP is only supported.
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Masahisa Kojima [Fri, 10 Nov 2023 04:25:40 +0000 (13:25 +0900)]
efi_loader: support boot from URI device path
This supports to boot from the URI device path.
When user selects the URI device path, bootmgr downloads
the file using wget into the address specified by loadaddr
env variable.
If the file is .iso or .img file, mount the image with blkmap
then try to boot with the default file(e.g. EFI/BOOT/BOOTAA64.EFI).
Since boot option indicating the default file is automatically
created when new disk is detected, system can boot by selecting
the automatically created blkmap boot option.
If the file is PE-COFF file, load and start the downloaded file.
The buffer used to download the ISO image file must be
reserved to avoid the unintended access to the image and
expose the ramdisk to the OS.
For PE-COFF file case, this memory reservation is done
in LoadImage Boot Service.
[Ilias fix a few memory leaks by replacing returns with gotos]
Lore: https://lore.kernel.org/u-boot/20231110042542.3797301-1-masahisa.kojima@linaro.org/T/#mbac31da301ff465b60894b38f3a587b2868cf817 Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org> Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Masahisa Kojima [Fri, 10 Nov 2023 04:25:39 +0000 (13:25 +0900)]
efi_loader: add return to efibootmgr event group
When the image loaded by efibootmgr returns, efibootmgr
needs to clean the resources. Adding the event of returning
to efibootmgr is useful to simplify the implementation.
Masahisa Kojima [Fri, 10 Nov 2023 04:25:38 +0000 (13:25 +0900)]
efi_loader: add missing const classifier for event service
const classifier is missing in EventGroup parameter of
CreateEventEx(). Fix it to remove the compiler warning.
NotifyContext parameter of CreateEventEx() is also defined
with const in UEFI specification, but NotifyContext parameter
of CreateEvent() is defined without const.
Since current implementation calls the common efi_create_event()
function from both CreateEventEx() and CreateEvent() services,
NotifyContext parameter leaves as is.
Raymond Mao [Fri, 10 Nov 2023 04:25:37 +0000 (13:25 +0900)]
efi_loader: Boot var automatic management
Changes for complying to EFI spec ยง3.5.1.1
'Removable Media Boot Behavior'.
Boot variables can be automatically generated during a removable
media is probed. At the same time, unused boot variables will be
detected and removed.
Please note that currently the function 'efi_disk_remove' has no
ability to distinguish below two scenarios
a) Unplugging of a removable media under U-Boot
b) U-Boot exiting and booting an OS
Thus currently the boot variables management is not added into
'efi_disk_remove' to avoid boot options being added/erased
repeatedly under scenario b) during power cycles
See TODO comments under function 'efi_disk_remove' for more details
The original efi_secboot tests expect that BootOrder EFI variable
is not defined. With this commit, the BootOrder EFI variable is
automatically added when the disk is detected. The original
efi_secboot tests end up with unexpected failure.
The efi_secboot tests need to be modified to explicitly set
the BootOrder EFI variable.
squashfs and erofs ls tests are also affected by this modification,
need to clear the previous state before squashfs ls test starts.
Co-developed-by: Masahisa Kojima <masahisa.kojima@linaro.org> Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org> Signed-off-by: Raymond Mao <raymond.mao@linaro.org> Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Joao Marcos Costa <jmcosta944@gmail.com> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Masahisa Kojima [Fri, 10 Nov 2023 04:25:35 +0000 (13:25 +0900)]
net: wget: add wget with dns utility function
Current wget takes the target uri in this format:
"<http server ip>:<file path>" e.g.) 192.168.1.1:/bar
The http server ip address must be resolved before
calling wget.
This commit adds the utility function runs wget with dhs.
User can call wget with the uri like "http://foo/bar".
Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org> Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Reviewed-by: Ramon Fried <rfried.dev@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Chris Packham [Fri, 27 Oct 2023 00:23:54 +0000 (13:23 +1300)]
Revert "arm64: Use FEAT_HAFDBS to track dirty pages when available"
This reverts commit 6cdf6b7a340db4ddd008516181de7e08e3f8c213. This is
part of a series trying to make use of the arm64 hardware features for
tracking dirty pages. Unfortunately this series causes problems for the
AC5/AC5X SoCs. Having exhausted other options the consensus seems to be
reverting this series is the best course of action.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Chris Packham [Fri, 27 Oct 2023 00:23:53 +0000 (13:23 +1300)]
Revert "arm64: Use level-2 for largest block mappings when FEAT_HAFDBS is present"
This reverts commit 836b8d4b205d2175b57cb9ef271e638b0c116e89. This is
part of a series trying to make use of the arm64 hardware features for
tracking dirty pages. Unfortunately this series causes problems for the
AC5/AC5X SoCs. Having exhausted other options the consensus seems to be
reverting this series is the best course of action.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Chris Packham [Fri, 27 Oct 2023 00:23:52 +0000 (13:23 +1300)]
Revert "armv8: enable HAFDBS for other ELx when FEAT_HAFDBS is present"
This reverts commit c1da6fdb5c239b432440721772d993e63cfdeb20. This is
part of a series trying to make use of the arm64 hardware features for
tracking dirty pages. Unfortunately this series causes problems for the
AC5/AC5X SoCs. Having exhausted other options the consensus seems to be
reverting this series is the best course of action.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Marcel Ziswiler [Thu, 26 Oct 2023 07:32:20 +0000 (09:32 +0200)]
imx: spl_imx_romapi: fix emmc fast boot mode case
This fixes a regression in the eMMC fast boot mode case where the buffer
was missing 464 bytes.
The code figures out how many bytes must at least be fetched to honor
the current read, rounds that up to the ss->pagesize [which is a no-op
in the USB download case because that has ->pagesize==1], fetches that
many bytes, but then recorded the original upper bound as the new end of
the valid data. However, this did not take into account the rounding up
to the ss->pagesize. Fix this by recording the actual bytes downloaded.
Fixes: 4b4472438f5a ("imx: spl_imx_romapi: avoid tricky use of spl_load_simple_fit() to get full FIT size") Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Acked-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk> Reviewed-by: Fabio Estevam <festevam@gmail.com>
John Keeping [Tue, 14 Nov 2023 11:30:17 +0000 (11:30 +0000)]
spl: fix TPL_SYS_MALLOC_F description
This config option enables the malloc() pool in TPL not the SPL. Fix
the description to accurately reflect this.
Fixes: fd8497dae54 (spl: Create proper symbols for enabling the malloc() pool) Signed-off-by: John Keeping <jkeeping@inmusicbrands.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Andre Przywara [Tue, 7 Nov 2023 16:09:00 +0000 (16:09 +0000)]
virtio: rng: gracefully handle 0 byte returns
According to the virtio v1.x "entropy device" specification, a virtio-rng
device is supposed to always return at least one byte of entropy.
However the virtio v0.9 spec does not mention such a requirement.
The Arm Fixed Virtual Platform (FVP) implementation of virtio-rng always
returns 8 bytes less of entropy than requested. If 8 bytes or less are
requested, it will return 0 bytes.
This behaviour makes U-Boot's virtio_rng_read() implementation go into an
endless loop, hanging the system.
Work around this problem by always requesting 8 bytes more than needed,
but only if a previous call to virtqueue_get_buf() returned 0 bytes.
This should never trigger on a v1.x spec compliant implementation, but
fixes the hang on the Arm FVP.
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Reported-by: Peter Hoyes <peter.hoyes@arm.com>
Simon Glass [Thu, 16 Nov 2023 01:35:23 +0000 (18:35 -0700)]
bootstd: Avoid freeing a non-allocated buffer
EFI applications can be very large and thus used to cause boot failures
when malloc() space was exhausted.
A recent changed fixed this by using the kernel_addr_r environment var
as the address of the buffer. However, it still frees the buffer when
the bootflow is discarded.
Fix this by introducing a flag to indicate whether the buffer was
allocated, or not.
Note that kernel_addr_r is not the last word here. It might be better
to use lmb to place images. But there is a lot of refactoring to do
before we can remove the environment variables. The distro scripts rely
on them so it is safe for bootstd to do so too.
Fixes: 6a8c2f9781c bootstd: Avoid allocating memory for the EFI file Signed-off-by: Simon Glass <sjg@chromium.org>
Reported by: Simon Glass <sjg@chromium.org>
Reported by: Shantur Rathore <i@shantur.com> Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Tested-by: Shantur Rathore <i@shantur.com>
Tom Rini [Fri, 17 Nov 2023 13:38:05 +0000 (08:38 -0500)]
Merge branch '2023-11-16-assorted-updates' into next
- squashfs improvements, remove common.h in some places, assorted code
fixes, fix a few CONFIG symbol names in Kconfig files, bring in
linux's <linux/time.h> conversion functions, poplar updates, bcb
improvements.
Sam Edwards [Sat, 11 Nov 2023 15:19:04 +0000 (08:19 -0700)]
fs: btrfs: fix reading when length specified
The btrfs read function limits the read length to ensure that it
and the read offset do not together exceed the size of the file.
However, this size was only being queried if the read length was
passed a value of zero (meaning "whole file"), and the size is
defaulted to 0 otherwise. This means the clamp will just zero out
the length if one is specified, preventing reading of the file.
Fix this by checking the file size unconditionally, and unifying
the default length and clamping logic as a single range check instead.
This bug was discovered when trying to boot Linux with initrd= via
'bootefi' from a btrfs partition. The EFI stub entered an infinite
loop of zero-length reads while trying to read the initrd, and the
boot process stalled indefinitely.
Signed-off-by: Sam Edwards <CFSworks@gmail.com> Reviewed-by: Qu Wenruo <wqu@suse.com>
Dmitrii Merkurev [Fri, 10 Nov 2023 05:59:55 +0000 (05:59 +0000)]
cmd: bcb: extend BCB C API to allow read/write the fields
Currently BCB C API only allows to modify 'command' BCB field.
Extend it so that we can also read and modify all the available
BCB fields (command, status, recovery, stage).
Co-developed-by: Cody Schuffelen <schuffelen@google.com> Signed-off-by: Cody Schuffelen <schuffelen@google.com> Signed-off-by: Dmitrii Merkurev <dimorinny@google.com> Cc: Eugeniu Rosca <erosca@de.adit-jv.com> Cc: Ying-Chun Liu (PaulLiu) <paul.liu@linaro.org> Cc: Simon Glass <sjg@chromium.org> Cc: Mattijs Korpershoek <mkorpershoek@baylibre.com> Cc: Sean Anderson <sean.anderson@seco.com> Cc: Cody Schuffelen <schuffelen@google.com> Tested-by: Mattijs Korpershoek <mkorpershoek@baylibre.com> # on vim3 Reviewed-by: Mattijs Korpershoek <mkorpershoek@baylibre.com>
Igor Opaniuk [Thu, 9 Nov 2023 22:34:34 +0000 (23:34 +0100)]
poplar: use random mac address
Set CONFIG_NET_RANDOM_ETHADDR=y, which sets random eth address in
case there is no configuration provided neither in CONFIG_ETHADDR
nor in "ethaddr" env variable.
This fixes the problem:
poplar# dhcp
Error: ethernet@9841000 address not set.
Signed-off-by: Igor Opaniuk <igor.opaniuk@foundries.io> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>
Igor Prusov [Thu, 9 Nov 2023 17:10:02 +0000 (20:10 +0300)]
linux/time.h: Add Linux time conversion defines
Currently there are no defines for time conversion in time.h, which
leads to drivers declaring those locally or not using defines at all, so
add them from Linux.
Signed-off-by: Igor Prusov <ivprusov@salutedevices.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Paul Barker [Wed, 8 Nov 2023 08:51:10 +0000 (08:51 +0000)]
sysreset: Fix unsupported request return values
The description of the sysreset request method in <sysreset.h> says that
the return value should be -EPROTONOSUPPORT if the requested reset type
is not supported by this device.
Signed-off-by: Paul Barker <paul.barker.ct@bp.renesas.com> Reviewed-by: Simon Glass <sjg@chromium.org>
This patch removes a number of struct and macro declaration that
were found through `git-grep` to be unused. Most of those are
related to compressor options and super block flags.
For reading a SquashFS image, we do not need the compressor options
or the flags. Those only encode settings used for packing the image,
mksquashfs uses them when appending data to an existing image. The
kernel implementation does not touch those, and we don't need them
either.
Signed-off-by: David Oberhollenzer <goliath@infraroot.at>
Sean Anderson [Sat, 28 Oct 2023 22:57:27 +0000 (18:57 -0400)]
test: eth: Don't crash if env_get returns NULL
env_get can return NULL if it fails to find the variable. Check its result
before using it.
Fixes: 6d9764c2a87 ("dm: test: Add a new test case against dm eth codes for NULL pointer access") Fixes: df33fd28897 ("test: eth: Add test for ethernet addresses") Signed-off-by: Sean Anderson <seanga2@gmail.com>
Tom Rini [Thu, 16 Nov 2023 17:55:48 +0000 (12:55 -0500)]
Merge patch series "spl: Use common function for loading/parsing images"
To quote the author:
This series adds support for loading all image types (Legacy (with and without
LZMA), FIT (with and without LOAD_FIT_FULL), and i.MX) to the MMC, SPI, NOR,
NET, FAT, EXT, NVMe, and semihosting load methods. It does this by introducing a
helper function which handles the minutiae of invoking the proper parsing
function, and reading the rest of the image.
Hopefully, this will make it easier for load methods to support all image types
that U-Boot supports, without having undocumented unsupported image types. I
applied this to several loaders which were invoking spl_load_simple_fit and/or
spl_parse_image_header, but I did not use it with others (e.g. DFU/RAM) which
had complications in the mix.
This series is organized roughly into two parts. Patches up to "spl: Add generic
spl_load function" are all setup or size-reduction oriented. Later patches
generally convert various load functions to spl_load.
bloat-o-meter results (for CONFIG_SPL only) at [1]. Size growth has been the
bigegst challenge to preparing this series. I have used every trick I can think
of to reduce bloat. Some SAMA boards no longer fit, but I have a plan to fix
them [2].
This is bar far the largest and most-difficult revision of this series to-date.
There are probably still things which can reduce the size, but I have been
working on this series for the better part of two months and I think it is a
good idea to get some feedback. Because of the SAMA bloat, this series will not
pass CI, so I expect to do a v7 before this is ready to apply. Feel free,
however, to apply patches in the first half (especially the fixes).
This version of the series is better-tested than ever before, thanks to some new
unit tests. However, things like the i.MX ROMAPI are untested. NAND should also
be tested more-widely, for reasons listed in the commit message. I encourage you
try this series out on your favorite board.
[1] https://gist.github.com/Forty-Bot/5bfe88676dd3c2aec6ebc23abb08e06f
This includes some changes to am335x_evm_spiboot and am65x_evm_r5_usbdfu
which have since been undone. This was ran for v6.
[2] https://lore.kernel.org/u-boot/20231105022742.632175-1-seanga2@gmail.com/
Sean Anderson [Wed, 8 Nov 2023 16:48:58 +0000 (11:48 -0500)]
spl: fat: Add option to disable DMA alignment
If we don't DMA-align buffers we pass to FAT, it will align them itself.
This behaviour likely should be deprecated in favor of
CONFIG_BOUNCE_BUFFER, but that's a task for another series. For the
meantime, don't bother aligning the buffer unless we had been doing so in
the past.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:57 +0000 (11:48 -0500)]
spl: spi: Consolidate spi_load_image_os into spl_spi_load_image
spi_load_image_os performs almost the same steps as the non-falcon-boot
path of spl_spi_load_image. The load address is different, and it also
loads a device tree, but that's it. Refactor the boot process so that
they can both use the same load function.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:56 +0000 (11:48 -0500)]
spl: Convert spi to spl_load
This converts the spi load method to use spl_load. The address used for
LOAD_FIT_FULL may be different, but there are no in-tree users of that
config. Since payload_offs is only used without OS_BOOT, we defer its
initialization.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:53 +0000 (11:48 -0500)]
spl: Convert nor to spl_load
This converts the nor load method to use spl_load. As a result it also
adds support for LOAD_FIT_FULL. Since this is the last caller of
spl_load_legacy_img, it has been removed.
We can't load FITs with external data with SPL_LOAD_FIT_FULL, so disable the
test in that case. No boards enable SPL_NOR_SUPPORT and SPL_LOAD_FIT_FULL, so
this is not a regression.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:51 +0000 (11:48 -0500)]
spl: Convert nand to spl_load
This converts the nand load method to use spl_load. nand_page_size may not
be valid until after nand_spl_load_image is called (see e.g. fsl_ifc_spl),
so we set bl_len in spl_nand_read. Since spl_load reads the header for us,
we can remove that argument from spl_nand_load_element.
There are two possible regressions which could result from this commit.
First, we ask for a negative address from spl_get_load_buffer. That is,
instead of
this could cause a problem if spl_get_load_buffer does not return valid
memory for negative offsets. Second, we now set bl_len for legacy images.
This can cause memory up to a bl_len - 1 before the image load address to
be written, which might not have been the case before. If this turns out to
be a problem, we can add an option for a bounce buffer.
We can't load FITs with external data with SPL_LOAD_FIT_FULL, so disable the
test in that case. No boards enable SPL_NAND_SUPPORT and SPL_LOAD_FIT_FULL, so
this is not a regression.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:50 +0000 (11:48 -0500)]
spl: Convert mmc to spl_load
This converts the mmc loader to spl_load. Legacy images are handled by
spl_load (via spl_parse_image_header), so mmc_load_legacy can be
omitted. To accurately determine whether mmc_load_image_raw_sector is used
(which might not be the case if SYS_MMCSD_FS_BOOT is enabled), we introduce
a helper config SYS_MMCSD_RAW_MODE. This ensures we can inline spl_load
correctly when a board only boots from filesystems. We still need to check
for SPL_MMC, since some boards enable configure raw mode even without MMC
support.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:47 +0000 (11:48 -0500)]
spl: Add generic spl_load function
Implementers of SPL_LOAD_IMAGE_METHOD have to correctly determine what
type of image is being loaded and then call the appropriate image load
function correctly. This is tricky, because some image load functions
expect the whole image to already be loaded (CONFIG_SPL_LOAD_FIT_FULL),
some will load the image automatically using spl_load_info.read()
(CONFIG_SPL_LOAD_FIT/CONFIG_SPL_LOAD_IMX_CONTAINER), and some just parse
the header and expect the caller to do the actual loading afterwards
(legacy/raw images). Load methods often only support a subset of the
above methods, meaning that not all image types can be used with all
load methods. Further, the code to invoke these functions is
duplicated between different load functions.
To address this problem, this commit introduces a "spl_load" function.
It aims to handle image detection and correct invocation of each of the
parse/load functions.
Although this function generally results in a size reduction with
several users, it tends to bloat boards with only a single user.
This is generally because programmers open-coding the contents of this
function can make optimizations based on the specific loader. For
example, NOR flash is memory-mapped, so it never bothers calling
load->read. The compiler can't really make these optimizations across
translation units. LTO solves this, but it is only available on some
arches. To address this, perform "pseudo-LTO" by inlining spl_load when
there are one or fewer users. At the moment, there are no users, so
define SPL_LOAD_USERS to be 0.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:46 +0000 (11:48 -0500)]
test: spl: Support testing LEGACY_LZMA filesystem images
These will soon be supported, so we need to be able to test it. Export the
lzma data and generally use the same process in spl_test_mmc_fs as
do_spl_test_load. If we end up needing this in third place in the future,
it would probably be good to refactor things out.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:44 +0000 (11:48 -0500)]
spl: nand: Remove spl_nand_legacy_read
Now that spl_nand_fit_read works in units of bytes, it can be combined with
spl_nand_legacy_read. Rename the resulting function spl_nand_read, since it
is no longer FIT-specific.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:43 +0000 (11:48 -0500)]
spl: Only support bl_len when we have to
Aligning addresses and sizes causes overhead which is unnecessary when we
are not loading from block devices. Remove bl_len when it is not needed.
For example, on iot2050 we save 144 bytes with this patch (once the rest of
this series is applied):
add/remove: 0/0 grow/shrink: 0/3 up/down: 0/-144 (-144)
Function old new delta
spl_load_simple_fit 920 904 -16
load_simple_fit 496 444 -52
spl_spi_load_image 384 308 -76
Total: Before=87431, After=87287, chg -0.16%
We use panic() instead of BUILD_BUG_ON in spl_set_bl_len because we still
need to be able to compile it for things like mmc_load_image_raw_sector,
even if that function will not be used.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:42 +0000 (11:48 -0500)]
spl: Remove filename from spl_load_info
For filesystems, filename serves the same purpose as priv. However,
spl_load_fit_image also uses it to determine whether to use a DMA-aligned
buffer. This is beneficial for FAT, which uses a bounce-buffer if the
destination is not DMA-aligned. However, this is unnecessary now that
filesystems set bl_len to ARCH_DMA_MINALIGN instead. With this done, we can
remove filename entirely.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:41 +0000 (11:48 -0500)]
spl: Set FAT bl_len to ARCH_DMA_MINALIGN
Instead of relying on the presence of filename to determine whether we are
dealing with a FAT filesystem (and should DMA-align the buffer), have FAT set
bl_len to ARCH_DMA_MINALIGN instead. With this done, we can remove the
special-case logic checking for the presence of filename.
Because filesystems are not block-based, we may read less than the size passed
to spl_load_info.read. This can happen if the file size is not DMA-aligned. This
is fine as long as we read the amount we originally wanted to. Modify the
conditions for callers of spl_load_info.read to check against the original,
unaligned size to avoid failing spuriously.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:40 +0000 (11:48 -0500)]
spl: Refactor spl_load_info->read to use units of bytes
Simplify things a bit for callers of spl_load_info->read by refactoring it
to use units of bytes instead of bl_len. This generally simplifies the
logic, as MMC is the only loader which actually works in sectors. It will
also allow further refactoring to remove the special-case handling of
filename. spl_load_legacy_img already works in units of bytes (oops) so it
doesn't need to be changed.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Tom Rini [Thu, 16 Nov 2023 17:46:09 +0000 (12:46 -0500)]
Merge patch series "nand: Add sandbox tests"
To quote the author:
This series tests raw nand flash in sandbox and fixes various bugs discovered in
the process. I've tried to do things in a contemporary manner, avoiding the
(numerous) variations present on only a few boards. The test is pretty minimal.
Future work could test the rest of the nand API as well as the MTD API.
Bloat (for v1) at [1] (for boards with SPL_NAND_SUPPORT enabled). Almost
everything grows by a few bytes due to nand_page_size. A few boards grow more,
mostly those using nand_spl_loaders.c. CI at [2].
Sean Anderson [Wed, 8 Nov 2023 16:48:38 +0000 (11:48 -0500)]
spl: Remove dev from spl_load_info
dev and priv serve the same purpose, and are never set at the same time.
Remove dev and convert all users to priv. While we're at it, reorder bl_len
to be last for better alignment.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:37 +0000 (11:48 -0500)]
spl: Remove NULL assignments in spl_load_info
Remove NULL assignments to fields in spl_load_info when .load doesn't
reference these fields. This can result in more efficient code. filename
must stay even if it is unused, since load_simple_fit uses it.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:36 +0000 (11:48 -0500)]
spl: semihosting: Don't close fd before spl_load_simple_fit
On real hardware, semihosting calls tend to have a large constant
overhead (on the order of tens of milliseconds). Reduce the number of
calls by one by reusing the existing fd in smh_fit_read, and closing it
at the end of spl_smh_load_image.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Wed, 8 Nov 2023 16:48:34 +0000 (11:48 -0500)]
arm: Disable SPL_FS_FAT when it isn't used
Several boards enable SPL_FS_FAT and SPL_LIBDISK_SUPPORT when they can't be
used (as there is no block device support enabled). Disable these configs.
The list of boards was generated with the following command:
LIBDISK was left enabled for the am* boards, since it seems to result in
actual size reduction, indicating that partitions are being used for
something.
Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Michal Simek <michal.simek@amd.com>
Sean Anderson [Wed, 8 Nov 2023 16:48:33 +0000 (11:48 -0500)]
spl: blk_fs: Fix uninitialized return value when we can't get a blk_desc
Initialize ret to avoid returning garbage if blk_get_devnum_by_uclass_id
fails.
Fixes: 8ce6a2e1757 ("spl: blk: Support loading images from fs") Signed-off-by: Sean Anderson <seanga2@gmail.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Sean Anderson [Sun, 5 Nov 2023 02:27:42 +0000 (22:27 -0400)]
arm: Enable SYS_THUMB_BUILD on AT91
Several AT91 boards are quite close to their SPL size limit. For example,
sama5d27_wlsom1_ek_mmc is just 173 bytes short of its limit and doesn't
even fit with older GCCs.
All AT91 processors should have thumb support. Enable SYS_THUMB_BUILD. This
shrinks SPL by around 30%.
Sean Anderson [Sat, 4 Nov 2023 20:37:53 +0000 (16:37 -0400)]
test: spl: Add a test for NAND
Add a SPL test for the NAND load method. We use some different functions to
do the writing from the main test since things like nand_write_skip_bad
aren't available in SPL.
We disable BBT scanning, since scan_bbt is only populated when not in SPL.
We use nand_spl_loaders.c as it seems to be common to at least a few boards
already. However, we do not use nand_spl_simple.c because it would require
us to implement cmd_ctrl. The various nand load functions are adapted from
omap_gpmc. However, they have been modified for simplicity/correctness.
Sean Anderson [Sat, 4 Nov 2023 20:37:52 +0000 (16:37 -0400)]
nand: Add sandbox driver
Add a sandbox NAND flash driver to facilitate testing. This driver supports
any number of devices, each using a single chip-select. The OOB data is
stored in-band, with the separation enforced through the API.
For now, create two devices to test with. The first is a very small device
with basic ECC. The second is an 8G device (chosen to be larger than 32
bits). It uses ONFI, with the values copied from the datasheet. It also
doesn't need too strong ECC, which speeds things up.
Although the nand subsystem determines the parameters of a chip based on
the ID, the driver itself requires devicetree properties for each
parameter. We do not derive parameters from the ID because parsing the ID
is non-trivial. We do not just use the parameters that the nand subsystem
has calculated since that is something we should be testing. An exception
is made for the ECC layout, since that is difficult to encode in the device
tree and is not a property of the device itself.
Despite using file I/O to access the backing data, we do not support using
external files. In my experience, these are unnecessary for testing since
tests can generally be written to write their expected data beforehand.
Additionally, we would need to store the "programmed" information somewhere
(complicating the format and the programming process) or try to detect
whether block are erased at runtime (degrading probe speeds).
Information about whether each page has been programmed is stored in an
in-memory buffer. To simplify the implementation, we only support a single
program per erase. While this is accurate for many larger flashes, some
smaller flashes (512 byte) support multiple programs and/or subpage
programs. Support for this could be added later as I believe some
filesystems expect this.
To test ECC, we support error-injection. Surprisingly, only ECC bytes in
the OOB area are protected, even though all bytes are equally susceptible
to error. Because of this, we take care to only corrupt ECC bytes.
Similarly, because ECC covers "steps" and not the whole page, we must take
care to corrupt data in the same way.
Sean Anderson [Sat, 4 Nov 2023 20:37:51 +0000 (16:37 -0400)]
arch: sandbox: Add function to create temporary files
When working with sparse data buffers that may be larger than the address
space, it is convenient to work with files instead. Add a function to create
temporary files of a certain size.
projects / u-boot.git / log