[GSoC] Debug Bootblock Stage for ARMv8 on QEMU

Hello again, I’m Asami. I’ve just finished 4 weeks as a GSoC student. I’m currently debugging the implementation of my main project, which is adding QEMU/AArch64 support. I can see nothing output right now when I start a QEMU with the coreboot.rom that has my implementation. It means there is something wrong before a hardware initialization has finished. In this article, I’m going to talk about what I found while debugging the bootblock for ARMv8.

Code Path of Bootblock Stage

The bootblock is executed just after CPU reset and it is almost written by assembly language. The main task is to set up a C-environment. The basic code path for ARMv8 from the beginning the bootblock to the romstage is:

  1. _start() at src/arch/arm64/armv8/bootblock.S
  2. arm64_init_cpu() at src/arch/arm64/armv8/cpu.S
  3. main() at srclib/bootblock.c
  4. run_romstage() at src/lib/prog_loaders.c
  5. prog_run() at src/lib/prog_ops.c
  6. arch_prog_run() at src/arch/arm64/boot.c
  7. main() at src/arch/arm64/romstage.c // The entry point of the romstage

You can use your custom _start function instead of the common _start function by CONFIG_BOOTBLOCK_CUSTOM=y and adding bootblock-y += bootblock_custom.S which is your custom assembly file.

The Reason Why Execution Stopped inside arm64_init_cpu()

I found that an execution stopped inside the arm64_init_cpu function for some reason. The line that has some problem is mrs x22, sctlr_el3 . MRS instruction can read a system control register and store the value into a general purpose register. So this line means to store the value of SCTLR_EL3 into the X22 register.

According to the “ARM Architecture Reference Manual ARMv8, for ARMv8-A architecture profile”, the purpose of SCTLR_EL3 is

Provides top level control of the system, including its memory system, at EL3. This register is part of the Other system control registers functional group.

Also, SCTLR_EL3 is accessible only from EL3 mode. EL3 is the highest privileged mode that a low-level firmware, including the secure monitor, works on it.

Next, I checked the current mode via mrs x0, CurrentEL. CurrentEL is a register that holds the current exception level. The result of CurrentEL was 0x04, which means the program works on EL1 mode. EL1 is the mode that an operating system kernel typically described as privileged. I didn’t have the right to access SCTLR_EL3. That’s why an execution stopped.

Ideas to Solve EL3 Issue

I considered 2 solutions:

  1. Use only EL1 registers
  2. Run QEMU in EL3

Firstly, I tried to use only EL1 registers. I replaced arm64_init_cpu with arm64_init_cpu_el1 that is a new function I created. Then I replaced SCTLR_EL3 with SCTLR_EL1 and TLBI ALLE3 with TLBI VMALLE1. It seems to work well but still, there was nothing output.

Secondly, I tried to run QEMU in EL3 that is enabled by -machine flag. QEMU can work on EL2 with -machine virtualization=on and EL3 with -machine secure=on to enable EL3. The following command works well for me.

$ qemu-system-aarch64 -bios ./build/coreboot.rom -M virt -cpu cortex-a53 -nographic -smp 1 -machine secure=on

My mentor Raul told me the second solution. Thank you so much!

[GSoC] Coreboot Coverity, week 4

Hello again! If you recall from my last post, the schedule this week is to fix the issues in northbridge/via and southbridge. However, Coverity is going through a major internal upgrade, and so the issue tracker has been offline all week. Luckily though I was able to fix most of these issues last week, so assuming the upgrade finishes soon I won’t be behind schedule. In the mean time, I decided to try flashing coreboot onto my T500, since the last component I was waiting for arrived last week. Here is a little mini-guide to my (sometimes harrowing) flashing experience.


  • ThinkPad T500
  • BeagleBone Black
  • 5V 2A power adapter for the BBB
  • Jumper Wires
  • Pomona 5252 Test Clip
  • Atheros AR9462 Wireless Card

Updating the EC

It is generally recommended to update the embedded controller firmware before flashing coreboot, which can only be done during a Lenovo BIOS update. (Unlike Chromebooks, ThinkPads unfortunately do not have open source EC’s.) I was able to find a copy of the latest BIOS on the Lenovo EOL Portal, and attempted to perform an update … which froze and crashed halfway through. Uh oh. This is OK, as long as I don’t restart the computer I can just try flashing it again, right? Wrong! The next time I tried it Windows ran into a fatal error and decided to force a restart for me (gah!). Upon booting it up again, I was met with absolutely nothing, because the screen wouldn’t even turn on. More than a little concerned that I had bricked it, I searched through online forums until I stumbled across the Crisis Recovery tool. Apparently, old ThinkPads have a method to force-update the BIOS from an external USB stick or floppy (if you have one of those lying around). The recovery tool had to be run in Windows XP Service Pack 3 emulation mode, and seemed to format the USB correctly. My ThinkPad wasn’t so impressed, and obstinately refused to recognize the stick. As a last hope, I asked around on IRC what to do, and Nico Huber informed me that the ThinkPad was likely not dead, and that I could just proceed with flashing coreboot anyway. Well, here goes nothing.

Building Coreboot

So we’re going to flash coreboot, but what options do I pick when compiling it? I scoured around the internet to find tutorials for flashing coreboot onto a T500 and other related ThinkPads, but they all recommended different options, sometimes contradictory. Hmmmm. Once again going back to IRC, Angel Pons helped me configure a very minimal build.

General setup ---> [*] Use CMOS for configuration values
              ---> [*] Allow use of binary-only repository
Mainboard ---> Mainboard vendor ---> Lenovo
          ---> Mainboard model ---> ThinkPad T500
Devices ---> Display ---> Linear "high-resolution" framebuffer

Now, the T500 is a very special laptop, in that it can run coreboot without any binary blobs at all. However, I decided to enable microcode updates anyway, since they provide important stability improvements (like not crashing). This laptop also comes with an Intel ME which can be completely wiped, but I decided to leave that for later. (Now that I know coreboot works, there will be a follow-up post in several weeks when I do that.)

Disassembly and Flashing

Like most laptops, the flash IC of the T500 is locked from the factory, and requires an initial external flash to install coreboot (afterwards, subsequent flashes can be done internally). This requires disassembling the laptop to access the SOIC-16, which is buried in the bowels of the T500 case and requires a complete tear-down to access. The Libreboot T500 page gives you a feel for the amount of work required to extract the motherboard, which along with the hardware maintenance manual I referred to extensively.

SOIC-16 highlighted in red

With the motherboard extracted from the case, the next step is to attach the Pomona 5252 to the SOIC-16 and jumper it to the BBB, which was all made very easy by this X200 guide. Somewhat blithely following the previous guide, I set up an old ATX PSU to provide 3.3v to the flash chip. However, whenever I connected it to the test clip, it would always power itself off. Strange. Going back to IRC, Nico informed me that this is in fact A VERY BAD AND DANGEROUS THING TO DO. THE INTERNET IS LYING – DO NOT USE AN ATX PSU, YOU COULD FRY YOUR MOTHERBOARD! Oops. After puzzling over how to provide enough power to the chip without the PSU, Patrick Rudolph chimed in that a) the T500 motherboard is basically indestructible (whew!), and b) the flasher itself should be able to provide enough power. Hooking the 3.3v cable into the BBB instead, I tried reading the flash chip.

$ flashrom -p linux_spi:dev=/dev/spidev1.0,spispeed=512
(a bunch of output that I forgot to write down)

It works! Even with a bricked Lenovo BIOS, it is still recommended to keep a backup, so next we read the old factory ROM.

$ flashrom -p linux_spi:dev=/dev/spidev1.0,spispeed=512 -r factory.rom

Do this three times with three distinct images, and compare their SHAsums to make sure they are all identical (otherwise the connection might be faulty). If they all match, keep one as a backup.

Now the moment of truth: writing coreboot.

$ flashrom -p linux_spi:dev=/dev/spidev1.0,spispeed=512 --ifd -i bios -w coreboot.rom

Note that because I left the ME as-is, it is important to only flash the BIOS region, not the entire chip.

Reassembly and Testing

Sadly, no instant gratification here – I had to reassemble half the laptop before I could test booting it up. However, after doing so and gingerly pressing the power button, I was greeted by the lovely SeaBIOS boot menu. It actually worked! Huzzah! Finishing reassembly, I replaced the factory Intel wireless card with an Atheros AR9462, which can run without any binary firmware. After installing Debian, I now have a laptop running completely free and open source software, all the way from the BIOS up (well, except for the ME, but I’ll fix that later).

T500 running coreboot and Debian 9

For the final icing on the cake, here is a fresh board status report for the T500. Many thanks to everyone who helped me in this process.

[GSoC] Common Mistakes for Beginners

Hello everyone. I am Asami and a student for this year’s GSoC project. My project is adding a new mainboard QEMU/AArch64 to make it easier for coreboot developers to support new boards for ARMv8. I’ve already written a small patch to enable building a sample program with libpayload for ARM architecture. Also, I’ve read the implementation of coreboot (main code path) for ARMv7 and QEMU (qemu/hw/arm/vexpress.c). Now, I just created a new CL for my main project and I started to read the implementation of the target machine of AArch64 (qemu/hw/arm/virt.c).

In this article, I’m going to talk about my mistakes when I developed coreboot. I hope it helps for beginners of coreboot development. The target board is QEMU/ARM and the CPU is ARMv7.

“ERROR: Ramstage region _postram_cbfs_cache overlapped by: fallback/payload”

I faced this error when I built coreboot.rom for QEMU/ARM with the coreinfo which is a small informational payload for coreboot. The cause is that the coreinfo doesn’t support ARM architecture and then the payload is compiled as a 32-bit x86.

Make sure that your payload is your target architecture. You need to use other executable files instead of the coreinfo when you want to use architectures other than x86. We provide the libpayload which is a small BSD-licensed static library.

The details of the error is:

$ make
W: Written area will abut bottom of target region: any unused space will keep its current contents
CBFS fallback/romstage
CBFS fallback/ramstage
CBFS config 
CBFS revision
CBFS fallback/payload
INFO: Performing operation on 'COREBOOT' region...
ERROR: Ramstage region _postram_cbfs_cache overlapped by: fallback/payload
Makefile.inc:1171: recipe for target 'check-ramstage-overlaps' failed
make: *** [check-ramstage-overlaps] Error 1 

“ERROR: undefined reference to ‘_ttb'” and “ERROR: undefined reference to ‘_ettb'”

This errors might happen when you build coreboot.rom by `make` at root directory. In this case, You need to add TTB() at your memleyout.ld.  

TTB is a translation table base address for MMU. TTBR0 and TTBR1 (TTB registers) hold the start point of TTB. We can put TTB anywhere in memory as long as we store the address to TTBR.

According to the “ARM Architecture Reference Manual ARMv7-A and ARMv7-R edition”, the difference between TTBR0 and TTBR1 is:

(B3-1345) When a PL1&0 stage 1 MMU is enabled, TTBR0 is always used. If TTBR1 is also used then:

– TTBR1 is used for the top part of the input address range

– TTBR0 is used for the bottom part of the input address range


(B4-1724) TTBCR determines which of the Translation Table Base Registers, TTBR0 or TTBR1, defines the base address for a translation table walk required for the stage 1 translation of a memory access from any mode other than Hyp mode.


TTBR0 is basically used for user processes and TTBR1 is used for kernel. However, Linux kernel only uses TTBR0 to reduce the time of context switch. (I just heard that Linux kernel starts to use TTBR1 because of security reasons such as Meltdown and Spectre.)

In coreboot, mmu_init() sets TTBR registers in arch/arm/armv7/mmu.c.

Fails to build a sample program with libpayload

We provide the libpayload which is a small BSD-licensed static library for coreboot and we also have a sample program to know how to use it. However, you might fail to build a sample program when you select the ARM architecture as a target with the following errors:

/usr/bin/ld: cannot represent machine `arm'

The reason why this problem happens is Makefile in the sample directory is old dated. So I created a CL to update current architectures that coreboot supports.

Please see the Makefile in https://review.coreboot.org/c/coreboot/+/33287

“Payload not loaded”

“Payload not loaded” happens when the load address of a payload is wrong. The load address should be placed in the RAM place where anyone can use. You can define the load address via CONFIG_LP_BASE_ADDRESS if you use a libpayload.

I created a CL for a sample configuration. Please see the config.emulation-qemu-arm in https://review.coreboot.org/c/coreboot/+/33287

Whole operations for building coreboot.rom with a sample payload for QEMU/ARM are:

1. Build a libc and cross compiler environment.

// In coreboot/payloads/libpayload/
$ make distclean // Always needs when switching a mainboard.
$ cp configs/config.emulation-qemu-arm configs/defconfig // Or you can set up it via 'make menuconfig'
$ make defconfig
$ make
$ make install

2. Build a sample payload hello.elf.

// In coreboot/payloads/libpayload/sample
$ make // Make sure that Makefile is updated by https://review.coreboot.org/c/coreboot/+/33287

3. Build coreboot.rom with a sample payload.

// In coreboot/
$ make distclean // Always needs when switching a mainboard.
$ make menuconfig // or make defconfig
  Select payload “payloads/libpayload/sample/hello.elf”
$ make

Make sure to do ‘make distclean’ before switching your board target

‘make distclean’ removes build artifacts and config files. The default archtecture in coreboot is x86, so you need to do ‘make distclean’ when you want to use other architectures.

Fails to update an existing CL on Gerrit 

Gerrit is a code review tool used in coreboot project. I’m familiar with GitHub and I thought the operations of Gerrit are almost the same with the operations of GitHub, but it weren’t.

On GitHub, developers can create a commit for each update. On the other hand, developers using Gerrit need to amend their commit until it will be merged.

Commands to create a new CL are almost the same with the operations of GitHub:

$ git add <target files>
$ git commit -s
$ git push

Commands to update an existing CL are slightly different:

$ git add <target files>
$ git commit --amend --no-edit

Make sure to leave the “Change-Id” line of the commit message as is.

[GSoC] Coreboot Coverity, Week 3

Hello again! This is a continuation of my posts about fixing the Coverity issues in coreboot. This week’s plan was to tackle the 28 issues in northbridge/intel, which turned out to be much easier than I expected, since I’m already done! With that out of the way, I’m going to begin working on northbridge/via and southbridge. For the curious, here is the project timeline for entire summer. (I had wanted to include this in last week’s post, but hadn’t figured out how to do tables in WordPress yet.)

Week Components Issues
May 6 to 10 util 22
May 13 to 17 util, payloads 22
May 20 to 24 arch, drivers 20
May 27 to 31 commonlib, cpu, lib, mainboard 22
June 3 to 7 northbridge/amd 21
June 10 to 14 northbridge/intel 28
June 17 to 21 northbridge/via, southbridge 22
June 24 to 28 soc/intel 21
July 1 to 5 soc/rockchip, soc/nvidia 20
July 15 to 19 soc/misc, vendorcode/cavium 26
July 22 to 26 vendorcode/amd 21
July 29 to Aug 2 vendorcode/amd 21
Aug 5 to 9 vendorcode/amd 20
Aug 12 to 16 vendorcode/amd 20
Aug 19 to 23 vendorcode/amd 20

As you can see, there are a lot of issues in the AMD vendorcode. This consists primarily of AGESA, AMD’s framework for initialization of their 64 bit platforms (somewhat similar to Intel’s FSP). This code is somewhat … dense (someone on IRC described it as a “sea of abstraction”), so I made sure to leave plenty of time for it. As always, you can keep up to date on my current progress on Gerrit.

PS: As an extra bonus, here is a picture of my new BeagleBone Black!

I recently got a ThinkPad T500 to practice installing coreboot on, and I needed some sort of external programmer to flash the SOIC. There are many options available (flashrom has a whole list here), but a single-board computer like this is one of the closest you can get to “plug-and-play.” There are many other popular boards (notably the Raspberry Pi), but the BBB doesn’t require any binary blobs to boot, and is open source hardware too. The only thing I’m waiting for now is an Atheros ath9k wireless card, which runs without any binary firmware. (Hey, if you’re gonna go freedom, you gotta go all the way.)

[GSoC] Coreboot Coverity, Introduction

Hello everyone! My name is Jacob Garber, and I am a student in this year’s GSoC 2019! My project is on making coreboot Coverity clean. Coverity is a free static-analysis tool for open source projects that searches for common coding mistakes and errors, such as buffer overruns, null pointer dereferences, and integer overflow. Coverity automatically analyzes the coreboot codebase and flags issues it finds, and my job is to classify them into bugs and false-positives and patch them if I can. You can check the Coverity overview for coreboot here, though seeing the issue tracker itself requires registration. At the beginning of the summer, coreboot had over 380 flagged issues, but it’s now down to 303, so we’re making progress! I plan to address 20-30 issues per week depending on the source component, which so far has gone surprisingly well (surprising, in the sense that coming into the summer I knew very little about coreboot or firmware development in general). For the curious, you can see the history and progress of all my changes on Gerrit. My mentors for this project are Patrick Georgi, Martin Roth, and David Hendricks, who have all been extremely helpful in guiding me through the development process, reviewing my patches, and answering my many questions. Thank you all.

Now, fixing Coverity bugs isn’t the only thing I’d like to do this summer. As I said before, I’d like to learn more about coreboot, and what better way to do that than installing it on a laptop! My current laptop is an old 2011 Macbook Air, which is surprisingly close to getting coreboot support (many thanks to Evgeny Zinoviev). However, I am (slightly) hesitant about installing yet-experimental firmware on my one and only development machine, so until then I picked up an old Thinkpad T500 to practice on. This laptop has the advantage of being able to run blob-free, and if in the very unlikely event I end up bricking it, who cares! (I mean, I’ll care, but it was a worthy sacrifice.) I also bought a BeagleBone Black to try out external flashing and was hoping to include a picture today, but the shipping was delayed. You’ll have to wait until next week!

[GSoC] Multiple status registers, block protection and OTP support, wrap-up (1/2)

Hello! 🙂

GSoC 2016 coding period has come to an end and mentor’s evaluating students this week. It has been an enriching 13 weeks of reading datasheets, designing structures, coding, learning and hanging out over IRC! 😛 I’d like to take this opportunity to present my work and details on how to use it. 🙂

Firstly, to offer context to the work, here is a list of public mails and blog posts. These should give an idea as to how the discussions and work evolved. A lot of the discussions have happened over IRC, but #flashrom does not keep any logs.

The patch sets that I sent to the mailing list can be found at –

  1. Multiple status register and access protection infrastructure
  2. OTP/Security registers infrastructure
  3. Dummy chips

You can also find these over at flashrom’s patchwork. The mailing list is where the review happens (although a better alternative, IMHO, is Gerrit which coreboot uses). The patches aren’t currently merged and are under review. In any case, you are most welcome to join review (which will likely be very helpful for me). 🙂 If you’d like to look at something more on the bleeding edge, then I invite you to my GitHub.

Now, moving on how to use the work. The most exhaustive documentation on how to use it is the code itself :P, but in the following list I attempt to list scenarios –

  • For SPI chips that have multiple status registers, flashrom’s verbose output will print the status register bits and there values. Most bits are named, i.e., the datasheet refers to the bit by an abbreviation, for instance, WEL for Write Enable Latch, WIP for Work In Progress, BP for Block Protect, LB for Lock Bit and so on. The verbose output will print these names, both in abbreviated and long forms, for most chips (and these abbreviations tend to be generic across many manufacturers). However, the process for adding new chips that leverage this, and adding new bits, is a fairly easy task (I would invite you to have a look at the code 😉 for more details). The verbose output also prints the write protection mode for status register(s) in effect (software protected, hardware protected, power cycle lock down and so on).
  • In case you want to disable or enable (a particular type) write protection for status register, you can use the --wp-disable or --wp-enable[=MODE] respectively (where MODE is either of software, hardware, power cycle or permanent – you are encouraged to have a look at the man page 🙂 for more details)
  • In case you want to protect a particular range of an SPI chip from writes or erases, you will need to alter the BP, TB or SEC bits. Currently, there is a CLI that will enable you to accomplish all that. 😛 First, you’ll want to look at the list of ranges your SPI chip supports – run flashrom with --wp-list. Take note of the start address and the length of the memory range you want to protect. Then again run flashrom with --wp-set-range start=0xfff000,len=4 (0xfff000 and 4 are for representational purpose only). By now the memory range is protected, but you can additionally enable status register write protection by following what the foregoing point described.
  • For SPI chips that support OTP, you can read, write and erase OTP regions (of course for supported chips :P). For OTP operations, you have at your disposal --print-otp-status, --read-otp [,reg=], --write-otp file=[,reg=], --erase-otp [reg=] and --lock-otp [reg=]. You can read the OTP memory to a file, or you can write to the OTP region from a file, very much like reading and writing from/to SPI chip. For more details, I would again like to point you to the man page. 🙂

Since this is a work-in-progress, the CLI may change (and is very likely). Currently around 10% of SPI chips use this new infrastructure. Models of a few manufacturers (and especially exotic ones like Atmel) are yet to be fully incorporated. You are most welcome to add support for new chips or update the existing ones to support new infrastructure. 🙂

I would like to sincerely thank my mentors Stefan and David for their support and help. I am indebted to them for this opportunity and I hope that we continue to share this relationship in the future while I continue to explore and contribute to flashrom. It has been a pleasure getting to know each of them. I’d also like to thank Urja for pitching in from time to time 🙂 It was fun hanging out over IRC and helping folks asking questions there. And I am looking forward to it for years to come. 😛

In the next and final part of this post, I will highlight how we intend to improve upon this work in the future, where it will be headed and what more we have in store, so please stay tuned. 😉 Phew, this was a long one, and rightly so as it attempts to summarise a great deal of experiences. If you have any feedback, questions or comments on the blogs or code, please feel to ping me on #flashrom where I am known as hatim. You can also email me at hatim@hatimak.me.

Thanks, and looking forward to hearing from you. 🙂 See you in the next and final part.

[GSoC] Better RISC-V support, wrap-up

In less than an hour, Google Summer of Code 2016 will be over (at least for us students). In this blog post, I will describe how you can use coreboot on RISC-V.

You can find the complete list of commits that I made during GSoC with this gerrit query.

The details

Compiling spike, the RISC-V instruction-set-level simulator

Spike, also known as riscv-isa-sim, is the reference implementation of RISC-V, and the only RISC-V platform that is currently known to work with coreboot (QEMU is nominally also supported, but the corresponding coreboot code has not been updated in a while).

First, you need to build and install libfesvr:

Then, you can compile and install spike:

  • Clone the spike git repository.
  • Apply the patch in this pull request to make console output possible.
  • Open riscv/processor.cc in a text editor and find processor_t::get_csr. Add a line that reads case CSR_MTIME: return 0;.
  • run ./configure --prefix="$HOME" --with-fesvr="$HOME" && make && make install

Compiling coreboot for RISC-V

  • Clone the coreboot git repository, if you haven’t already
  • Apply Iru Cai’s patch that updates the toolchain to GCC 6.1. You will currently have to fix a merge conflict when you apply this patch, but it’s an easy one.
  • Run make crossgcc-riscv
  • Run make menuconfig to configure coreboot. Select Emulation and SPIKE usb riscv in the Mainboard menu.
  • Run make
  • Run util/riscvtools/make-spike-elf.sh build/coreboot.rom build/coreboot.elf
  • Start coreboot by running spike build/coreboot.elf. You should see a few pages of output, ending in Payload not loaded.

Compiling and running Linux

  • Clone the riscv-linux git repository and check out the priv-1.9 branch
  • Apply this patch that allows linux to be started in machine-mode.
  • Download a copy linux 4.6.x from kernel.org, and unpack it. I’ll assume version 4.6.2 is used.
  • cd into linux-4.6.2/arch and symlink the arch/riscv directory from riscv-linux
  • Back in linux-4.6.2, run make O=build ARCH=riscv defconfig; cd into the newly created build directory.
  • Run make ARCH=riscv menuconfig. In the “General Setup” menu of the linux menuconfig, enter path/to/coreboot/util/crossgcc/xgcc/bin/riscv64-elf- as the cross-compiler tool prefix.
  • Run make ARCH=riscv vmlinux to compile linux.
  • Open vmlinux in a hex editor, such as hexer. Change the 8-byte number at 0x18 to 00 00 00 90 00 00 00 00; Add 00 00 00 90 00 00 00 00 to the numbers at 0x58 and 0x90, to load linux at physical addresses within RAM. It’s unfortunate that I have to recommend this step, but I did not come up with a better fix yet.

Next, you need to add vmlinux to coreboot:

  • Return to the coreboot directory.
  • Apply the remaining coreboot patches that are tagged riscv.
  • In the Payload menu, select An ELF executable payload. Instead of payload.elf, select the vmlinux file.
  • Run make and util/riscvtools/make-spike-elf.sh build/coreboot.rom build/coreboot.elf again.
  • Run spike build/coreboot.elf. You should now see a Linux boot, at least partially.

Future work

Even though my GSoC is over, coreboot’s support for RISC-V can still be improved, and I intend to fix at least some of the following things:

  • As you can see above, running coreboot and linux on RISC-V currently isn’t straight-forward, but involves a few manual steps.
  • There are other RISC-V platforms that I’d like to see coreboot on, such as lowRISC.
  • Linux does not completely boot, i.e. into userspace. There are still some bugs to be ironed out.
  • Automatic testing could be used to detect regressions.
  • I only tested coreboot on RISC-V with Linux; support for other operating systems or payloads is welcome.


I’d like to thank Ron Minnich and Furquan Shaikh for being good mentors, and everyone in the coreboot community for being helpful and friendly.

[GSoC] Better RISC-V Support, week #10

This past week, I worked on the virtual memory initialization code of coreboot on RISC-V. The first part of this was to update encoding.h a file that defines constants such as bit masks which are necessary for interacting with RISC-V’s Control and Status Registers. As a result, I also had to change a few files that relied on outdated constants. Then I wrote some code to walk the page table structures, and fixed one or two bugs in the page table setup code. Unfortunately my patches aren’t as finished as I would like them to be.

When I tested my changes with a Linux payload (which I had to edit the ELF headers of, because Linux uses virtual addresses, while cbfstool and the SELF loader use physical addresses), I stumbled upon another strange error: I get a store access fault in the instruction of the trap handler that saves the first register, even if I initialize the stack pointer to a value that’s within the RAM. When the trap handler runs again to handle this store access fault, it runs without any problems. This fault is especially confusing, because machine mode should always be able to access RAM through its physical addresses.

What’s next?

During the next week, I’ll be traveling and won’t be able to work on coreboot. When I return, I will rework my patches so they can be merged, and hopefully understand the aforementioned access fault problem. Properly set-up page tables should bring me a step closer to running Linux on coreboot/RISC-V (without bbl in the middle).

If time permits, I will start porting coreboot to the Nexys 4 DDR devboard.

[GSoC] Multiple status registers, block protection and OTP support, week #6, 7, 8 and 9

Hello! I have been away for some time now, so this is going to be a longer post. I hope you have missed hearing from me 😛 In this post I will talk about my work in the weeks post-midterm evaluations. After a discussion with my mentors in the midterm evaluations week, we decided to shift focus towards the first three phases of my GSoC proposal for the remainder of the duration. Work on the final phase will be done after GSoC along with the more long-term goals that have come up as I have been making progress.

I submitted patches (finally ;)) to the mailing list. The set of patches adds multiple status register, and block protection infrastructure. I have also added a command-line interface to expose the new functionality. Although I am not sure that the exact wording of the CLI is most optimum, but I did not spend a lot of time on that because IMO it is a rather subjective issue and altering it is not a difficult task. The set of patches also adds support for new infrastructure to around 90 existing chips. I am still waiting to receive feedback and review on them. (My mentors had been slightly busier then.) I am also investigating adding support for access protection to non-SPI chips. This isn’t on the highest priority (more like a long-term goal), but once the SPI infrastructure gets merged, I will begin writing code for that.

Based on the initial prototypes I built (here, here and here), we had decided to use pointers to new structs instead of fully embedding them in struct flashchip. This decision really started to show when I was adding support for existing chips – with only 25 unique struct definitions we were able to support those 90 chips! 😉 One of the problems I faced was that I needed to test the new infrastructure, but doing so on a physical chip would be cumbersome. So that problem was solved by adding a dummy chip to use with flashrom’s dummy programmer. (At that time the code was the dummy chip was messy and something I would be ashamed to put up publicly, but now I have improved upon it! :P)

Currently I am working on finalising the OTP/security register(s) patches – more specifically, adding support to existing chips, code cleanup and documentation. I will be able to send them to the mailing list in a few days. In my research on Eon, GigaDevice and Winbond chips, 2 distinct models for OTP were observed – the GigaDevice and Winbond model with security register(s), and the Eon model with a security sector.

The Security Register(s) model has 3 separate opcodes for read, program and erase – 0x48, 0x42 and 0x44 respectively. A chips can have multiple security registers (most commonly 3, but as high as 4) with each register being anywhere between 128 bytes to 1024 bytes in size (most commonly 512 bytes and then 256 bytes). Usually chips have a lock bit (LB1, LB2, …) in the status register that correspond to respective security registers. These one-time programmable bits are changed using the standard WRSR instruction. Some chips have a single lock bit that controls OTP status for all security registers.

The Security Sector model has a separate sector which can be operated in the OTP mode. OTP mode is entered with opcode 0x3A and exited by sending WRDI (0x04) instruction. While in the OTP mode, the sector behaves just like any other sector – normal read, program and erase instructions apply. The SRP/SRWD bit is served as OTP bit while in OTP mode. Issuing the WRSR command (irrespective of the data sent along with it) will cause the one-time programmable OTP bit to be set.

One of the recurrent issues (for the lack of a better word, I don’t think of it to be an issue really ;)) is that many chips I have based my research on, are not originally supported by flashrom (perhaps unfortunate siblings of the same family that didn’t find support in flashrom earlier xD). I don’t call it an issue per se because after I have submitted my patches flashrom will end up supporting even more chips, but since I have to write more code it might take slightly longer to submit the patches.

There is a third model which is dominantly followed by Spansion chips and a couple of AMIC chips (some AMIC chips follow one of the earlier models – it’s like AMIC couldn’t decide which one to stick to or they probably had different teams working on it! :P). Similar to the security sector design, these chips also have a separate OTP sector but instead of storing configuration in the register, a byte within the sector is allocated for storing the configuration data. I have planned to support this model in the next revision of patches, after the upcoming ones get reviewed and merged.

Thanks for your time, it was nice to get back in touch with you! 🙂
(Phew, that was long! :P)

[GSoC] Better RISC-V support, week #6/7/8/9

Since I haven’t posted an update of my coreboot-on-RISC-V work in a while, this will be a slightly longer post.

Week 6

In week 6, I started documenting how to build and boot coreboot on RISC-V, in the coreboot wiki.
It is now a bit outdated, because we’re moving away from using bbl to boot Linux.

Week 7

I wrote some patches: I removed code that used the old Host-Target Interface (HTIF), because it’s deprecated. I submitted an improved version of my workaround for the bug that causes Spike to only execute 5000 instructions in some cases. I informed the coreboot resource management subsystem about the position of the RAM in physical address space, so that the program loader wouldn’t refuse to load segments into RAM. I submitted two patches to fix compiler errors with the new toolchain.

Meanwhile, there were some good news in the RISC-V world:

Week 8

I submitted a few more patches and started to explore the Nexys4 board. The precompiled bitstream and kernel from the lowRISC version 0.3 tutorial worked without any problems, and after a few days and some help from the lowRISC mailing list, I was able to recompile the lowRISC bitstream.

This week

I discussed the choice of boot medium with the lowRISC developers, and they agreed that a memory-mapped flash would be useful. Once it is implemented, I
can start porting coreboot to lowRISC on the Nexys 4 DDR board. Luckily the Nexys4 already has large enough flash to use for this purpose.

My mentors and I agreed that the switch from machine mode to supervisor mode should be left completely to the payload.

What’s next?

I will continue to work on running coreboot on the Spike emulator. Currently I’m facing the following problems and tasks:

  • Linux, when compiled to an ELF file (vmlinux) specifies that it wants to be loaded at the physical address 0x0 and at the virtual address 0xffffffff80000000. Since coreboot’s ELF loading code only looks at the physical address, it refuses to load Linux, since RAM starts at 0x80000000 on RISC-V.
  • Low level platform information (most importantly the memory layout) is passed to the firmware (coreboot in this case) as a configuration string, which is dynamically generated by the emulator, in the case of Spike. I still need to implemented a parser for this format, so coreboot can know how much memory is available.
  • The RISC-V Privileged Architecture Specification 1.9 specifies that there shall be a page at the top of the virtual address space where the operating system can call a few functions exposed by the firmware (this is the Supervisor Binary Interface).