summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2018-04-06efi_loader: Check machine type in the image headerIvan Gorinov
Check FileHeader.Machine to make sure the EFI executable image is built for the same architecture. For example, 32-bit U-Boot on x86 will print an error message instead of loading an x86_64 image and crashing. Signed-off-by: Ivan Gorinov <ivan.gorinov@intel.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_loader: completely initialize networkHeinrich Schuchardt
Add missing network initialization code. Before the patch the network was only usable if a network command like dhcp or tftp had beed executed. This was visible when interrupting the console countdown and executing bootefi selftest for vexpress_ca15_tc2_defconfig. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05arm: print information about loaded UEFI imagesHeinrich Schuchardt
If an exception occurs in a UEFI loaded image we need the start address of the image to determine the relocation offset. This patch adds the necessary lines after the registers in the crash dump. A possible output would be: UEFI image [0xbffe6000:0xbffe631f] pc=0x138 '/\snp.efi' With the offset 0x138 we can now find the relevant instruction in the disassembled 'snp.efi' binary. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_loader: new functions to print loaded image informationHeinrich Schuchardt
Introduce functions to print information about loaded images. If we want to analyze an exception in an EFI image we need the offset between the PC and the start of the loaded image. With efi_print_image_info() we can print the necessary information for a single image, e.g. UEFI image [0xbffe6000:0xbffe631f] pc=0x138 '/\snp.efi' efi_print_image_infos() provides output for all loaded images. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05mmc: use core clock frequency in bcm2835 sdhostJonathan Gray
In raspberrypi-firmware 7fdcd00e00a42a1c91e8bd6f5eb8352fe9358557 and later start.elf now sets the EMMC clock to 200 MHz. According to Phil Elwell in https://github.com/raspberrypi/firmware/issues/953 the SDHost controller shares the core/VPU clock and doesn't use the EMMC clock. Use the core clock id when determining the frequency to allow U-Boot to work with recent versions of raspberrypi-firmware. Otherwise U-Boot hangs at: U-Boot 2018.03 (Mar 14 2018 - 20:36:00 +1100) DRAM: 948 MiB RPI 3 Model B (0xa02082) MMC: mmc@7e202000: 0, sdhci@7e300000: 1 Loading Environment from FAT... Signed-off-by: Jonathan Gray <jsg@jsg.id.au> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05rpi3_32b: Enable lan78xx driverPeter Robinson
The new Raspberry Pi B 3+ has a lan78xx device attached to it. Let's add driver support in U-Boot for it. Signed-off-by: Peter Robinson <pbrobinson@gmail.com> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05rpi: Add identifier for the new RPi3 B+Alexander Graf
The Raspberr Pi Foundation released a new RPi3 version which we want to detect as well, so we can enable ethernet on it and know the correct device tree file name. Add an identifier for it. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05rpi3: Enable lan78xx driverAlexander Graf
The new Raspberry Pi B 3+ has a lan78xx device attached to it. Let's add driver support in U-Boot for it. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05rpi: Allow to boot without serialAlexander Graf
When we enable CONFIG_OF_BOARD on Raspberry Pis, we may end up without serial console support in early boot. Hence we need to make the serial port optional, otherwise we will never get to the point where serial would be probed. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_selftest: test getinfo(EFI_FILE_SYSTEM_INFO)Heinrich Schuchardt
Check that the getinfo() service of the file protocol correctly returns the partion label. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_selftest: partition label for test imageHeinrich Schuchardt
efi_selftest_disk_image.h contains a disk image. We use it to test the EFI_FILE_PROTOCOL. The patch sets the partition label to 'U-BOOT TEST'. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_loader: implement EFI_FILE_SYSTEM_INFOHeinrich Schuchardt
Implement the information type EFI_FILE_SYSTEM_INFO in the service GetInfo() of the EFI_FILE_PROTOCOL. The volume label is not available in U-Boot. As a work-around use the partition name instead. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_loader: ascii2unicode(): add trailing \0Heinrich Schuchardt
When converting an ASCII string to UTF-16 don't forget to copy the trailing \0. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-05efi_loader: use const for GUIDs in the EFI_FILE_PROTOCOLHeinrich Schuchardt
Use const efi_guid_t* when passing GUIDs. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04Merge git://git.denx.de/u-boot-sunxiTom Rini
Signed-off-by: Tom Rini <trini@konsulko.com>
2018-04-04efi_loader: correctly determine the boot partitionHeinrich Schuchardt
The device path of the loaded image should be set to the partition from which the image was loaded. This requires using the same logic as the load command. Without the patch the device path pointed to the whole disk after executing load mmc 0: 0x43000000 FILE and not to the boot partition from which the file was actually loaded. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: use correct types in EFI_FILE_PROTOCOLHeinrich Schuchardt
In the EFI_FILE_PROTOCOL buffer sizes and positions are passed as UINTN and not as u64. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: correct types for EFI_LOADED_IMAGE_PROTOCOLHeinrich Schuchardt
We should not use void * but specific types for * device_handle * file_path Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: ImageSize must be multiple of SectionAlignmentHeinrich Schuchardt
According to the Portable Executable and Common Object File Format Specification the image size must be a multiple of the alignment of sections. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: save image relocation address and sizeHeinrich Schuchardt
For analyzing crash output the relocation address and size are needed. Save them in the loaded image info. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: use efi_uintn_t for LoadImageHeinrich Schuchardt
We generally use efi_uintn_t where the UEFI spec uses UINTN. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: initialize device path on allocPatrick Wildt
Since the backing memory for a new device path can contain stale data we have to make sure that we zero the buffer. Otherwise some code paths that don't set all fields in a structure backed by this device path might contain unwanted stale data. Signed-off-by: Patrick Wildt <patrick@blueri.se> Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: complete efi_pxe_mode struct definitionPatrick Wildt
The efi_pxe_mode struct which represents the PXE_BASE_CODE_PROTOCOL Replace the placeholder paddings in the efi_pxe_mode struct with the actual fields as defined in UEFI's PXE Base Code Protocol. Since our efi_ip_address is a simple char array set a specific alignment to the efi_ip_address fields, as expected by the UEFI spec. Signed-off-by: Patrick Wildt <patrick@blueri.se> [agraf: s/pxe_discovervalid/pxe_discover_valid] Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: Fix network DP with DM_ETHAlexander Graf
When CONFIG_DM_ETH is set, we assemble the device path properly with a full device hierarchy. Our helper function dp_fill() even put the MAC node itself in it for us. However, for non-DM compatibility we also have code in that added the MAC node manually. That code now runs on top of the existing MAC node: Handle 0x3db2f6b0 /HardwareVendor(e61d73b9-a384-4acc-aeab-82e828f3628b)[0: ] /USBClass(0,0,9,0,0)/USBClass(424,9514,9,0,2)/MacAddr(b8:27:eb:e1:81:47,1) /MacAddr(b8:27:eb:e1:81:47,57)/EndEntire We obviously don't need the additional node and in fact, grub chokes on it and fails to match the DP against the ethernet device node. So this patch moves the additional MAC node into the non-DM code path: Handle 0x3db3fde0 /HardwareVendor(e61d73b9-a384-4acc-aeab-82e828f3628b)[0: ] /USBClass(0,0,9,0,0)/USBClass(424,9514,9,0,2)/MacAddr(b8:27:eb:e1:81:47,1) /EndEntire While at it, we also mark the non-DM MAC node as ethernet. Fixes: b66c60dde9d ("efi_loader: add device-path utils") Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: Fix return value for efi_add_runtime_mmioAlexander Graf
The efi_add_runtime_mmio function incorrectly returned the added address as return value rather than EFI_SUCCESS. Fix it by checking the return value of efi_add_memory_map properly. Fixes: f057cfef5dc ("efi_loader: exit status for efi_reset_system_init") Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: Optimize GOP moreAlexander Graf
The GOP path was optimized, but still not as fast as it should be. Let's push it even further by trimming the hot path into simple 32bit load/store operations for buf->vid 32bpp operations. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: Optimize GOP switchAlexander Graf
We usually try to compile for size, not for speed. Unfortunately with the more powerful GOP infrastructure to handle all sorts of GOP operations, we end up slowing down our copying hot path quite a lot. So this patch moves the 4 possible GOP operation modes into separate functions which call a common function again. The end result of that is more optimized code that can properly do constant propagation throughout its switch() statements and thus removes compares in the hot path. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: use __always_inline for pixel conversionHeinrich Schuchardt
We optimize for size using -Os so gcc might ignore 'inline'. Pixel conversions are called so often that we always want to inline them. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: RestoreTPL should execute queued eventsHeinrich Schuchardt
When the TPL is lowered queued events may become eligible for execution. iPXE uses the following pattern to request event execution: bs->RestoreTPL ( TPL_APPLICATION ); bs->RaiseTPL ( TPL_CALLBACK ); Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: use TPL_NOTIFY for network timer eventHeinrich Schuchardt
We use a timer to poll the network. iPXE is used for booting from iSCSI drives. It has been changed to run at TPL_CALLBACK most of the time (which is not what the UEFI spec recommends). By changing our timer to TPL_NOTIFY we can ensure that it is nevertheless executed. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04git: mailrc: Add myself and efi to listAlexander Graf
I wasn't listed in the mailrc before, let's fix that. Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: correctly support parameter delta in BltHeinrich Schuchardt
In the Blt service of the EFI_GRAPHICS_OUTPUT_PROTOCOL the parameter delta is measured in bytes and not in pixels. The coding only supports delta being a multiple of four. The UEFI specification does not explicitly require this but as pixels have a size of four bytes we should be able to assume four byte alignment. The corresponding unit test is corrected, too. It can be launched with setenv efi_selftest block image transfer bootefi selftest Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_selftest: unit test for event groupsHeinrich Schuchardt
Supply a unit test for event groups. Create multiple events in an event group. Signal each event once and check that all events are notified once in each round. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_selftest: fix device tree unit testHeinrich Schuchardt
Include libfdt.h was moved by commit b08c8c487083 ("libfdt: move headers to <linux/libfdt.h> and <linux/libfdt_env.h>") Fixes: e236200c7fa6 ("efi_selftest: check installation of the device tree") Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi: Correct header order in efi_memorySimon Glass
The headers are not in the correct order. Fix this. Also drop libfdt_env.h since it is not needed. Signed-off-by: Simon Glass <sjg@chromium.org> Rebased Reviewed-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: use constants in efi_allocate_pages()Heinrich Schuchardt
Using the existing predefined constants is less error prone and makes the code easier to read. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: show UEFI revision in helloworldHeinrich Schuchardt
Output the UEFI revision number in helloworld. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_selftest: unit test for EFI_SIMPLE_TEXT_INPUT_PROTOCOLHeinrich Schuchardt
Provide a unit test for the EFI_SIMPLE_TEXT_INPUT_PROTOCOL. The unicode character and the scan code are printed for text input. To run the test: setenv efi_selftest text input bootefi selftest Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: add missing EFI_RESET_PLATFORM_SPECIFICHeinrich Schuchardt
EFI_RESET_PLATFORM_SPECIFIC is one of the values that can be used for the EFI service ResetSystem. The missing definition is added. The value has to handled in efi_reset_system(). Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_selftest: test gop bitbltHeinrich Schuchardt
The test checks all block image transfer operations of the graphical output protocol. An animated submarine is shown. To run the test: setenv efi_selftest bock image transfer bootefi selftest Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: implement missing bit blit operations in gopHeinrich Schuchardt
With the patch all block image transfer operations of the EFI_GRAPHICS_OUTPUT_PROTOCOL are supported. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: implement event groupsHeinrich Schuchardt
If an event of a group event is signaled all other events of the same group are signaled too. Function efi_signal_event is renamed to efi_queue_event. A new function efi_signal_event is introduced that checks if an event belongs to a group and than signals all events of the group. Event group notifciation is implemented for ExitBootServices, InstallConfigurationTable, and ResetSystem. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: define GUIDS for event groupsHeinrich Schuchardt
Event groups are used to signal multiple events at the same time. They are identified by GUIDs. This patch provided the predefined GUIDs of UEFI specification 2.7. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: manage events in a linked listHeinrich Schuchardt
Lift the limit on the number of events by using a linked list. This also allows to have events with type == 0. This patch is based on Rob's patch efi_loader: fix events https://lists.denx.de/pipermail/u-boot/2017-October/309348.html Suggested-by: Rob Clark <robdclark@gmail.com> Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: fix formatting errorsHeinrich Schuchardt
Fix formatting errors in efi_boottime.c indicated by scripts/checkpatch.py. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: clear signaled state in CheckEventHeinrich Schuchardt
CheckEvent must clear the signaled state. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: check parameter in InstallConfigurationTableHeinrich Schuchardt
Check that parameter guid is not NULL. This avoids a possible NULL pointer exception. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: correct input of special keysHeinrich Schuchardt
Don't set unicode_char if scan_code is set. Add support for page up, page down, and insert. Correct input of function keys. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_selftest: check installation of the device treeHeinrich Schuchardt
The unit test checks if a device tree is installed. It requires that the 'compatible' property of the root node exists. If available it prints the 'serial-number' property. The serial-number property is derived from the environment variable 'serial#'. This can be used to check if the image_setup_libfdt() function is executed. A Python test is supplied. It sets a value for serial# and checks that the selftest shows this as serial-number. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
2018-04-04efi_loader: support device tree for bootefi selftestHeinrich Schuchardt
The second argument of the bootefi command should always be usable to specify a device tree. This was missing for bootefi selftest and bootefi hello. Proper error handling is added. Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Alexander Graf <agraf@suse.de>
generated by cgit (git 2.34.1) at 2025-01-21 08:36:45 +0000