qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v2] tests/boot_linux_console: add a test for ris


From: Alistair Francis
Subject: Re: [Qemu-devel] [PATCH v2] tests/boot_linux_console: add a test for riscv64 + virt
Date: Wed, 17 Jul 2019 15:44:03 -0700

On Tue, Jul 16, 2019 at 10:21 PM Chih-Min Chao <address@hidden> wrote:
>
>
>
> On Wed, Jul 17, 2019 at 6:59 AM Alistair Francis <address@hidden> wrote:
>>
>> On Tue, Jul 16, 2019 at 2:50 PM Philippe Mathieu-Daudé
>> <address@hidden> wrote:
>> >
>> > On 7/16/19 10:43 PM, Alistair Francis wrote:
>> > > On Tue, Jul 16, 2019 at 6:56 AM Chih-Min Chao <address@hidden> wrote:
>> > >>
>> > >>
>> > >> On Tue, Jul 16, 2019 at 12:34 AM Alistair Francis <address@hidden> 
>> > >> wrote:
>> > >>>
>> > >>> On Mon, Jul 15, 2019 at 6:02 AM Philippe Mathieu-Daudé
>> > >>> <address@hidden> wrote:
>> > >>>>
>> > >>>> On 7/15/19 1:09 PM, Chih-Min Chao wrote:
>> > >>>>> On Mon, Jul 15, 2019 at 5:15 PM Philippe Mathieu-Daudé
>> > >>>>> <address@hidden <mailto:address@hidden>> wrote:
>> > >>>>>
>> > >>>>>     On 7/15/19 11:08 AM, Chih-Min Chao wrote:
>> > >>>>>     > Similar to the mips + malta test, it boots a Linux kernel on a 
>> > >>>>> virt
>> > >>>>>     > board and verify the serial is working.  Also, it relies on 
>> > >>>>> the serial
>> > >>>>>     > device set by the machine itself.
>> > >>>>>     >
>> > >>>>>     > If riscv64 is a target being built, "make check-acceptance" 
>> > >>>>> will
>> > >>>>>     > automatically include this test by the use of the 
>> > >>>>> "arch:riscv64" tags.
>> > >>>>>     >
>> > >>>>>     > Alternatively, this test can be run using:
>> > >>>>>     >
>> > >>>>>     >   $ avocado run -t arch:riscv64 tests/acceptance
>> > >>>>>     >
>> > >>>>>     > packages
>> > >>>>>     >   debian official
>> > >>>>>     >     binutils-riscv64-linux-gnu_2.32-8
>> > >>>>>     >     opensbi_0.4-1_all
>> > >>>>>     >     linux-image-4.19.0-5-riscv64 4.19.37-4
>> > >>>>>     >   third-party
>> > >>>>>     >
>> > >>>>>      
>> > >>>>> https://github.com/groeck/linux-build-test/rootfs/riscv64/rootfs.cpio.gz
>> > >>>>>     >     (the repo is also used in mips target acceptance)
>> > >>>>>     >
>> > >>>>>     > Signed-off-by: Chih-Min Chao <address@hidden
>> > >>>>>     <mailto:address@hidden>>
>> > >>>>>     > ---
>> > >>>>>     >  .travis.yml                            |  2 +-
>> > >>>>>     >  tests/acceptance/boot_linux_console.py | 66
>> > >>>>>     ++++++++++++++++++++++++++++++++++
>> > >>>>>     >  2 files changed, 67 insertions(+), 1 deletion(-)
>> > >>>>>     >
>> > >>>>>     > diff --git a/.travis.yml b/.travis.yml
>> > >>>>>     > index 5d3d6ee..21fcead 100644
>> > >>>>>     > --- a/.travis.yml
>> > >>>>>     > +++ b/.travis.yml
>> > >>>>>     > @@ -232,7 +232,7 @@ matrix:
>> > >>>>>     >
>> > >>>>>     >      # Acceptance (Functional) tests
>> > >>>>>     >      - env:
>> > >>>>>     > -        - CONFIG="--python=/usr/bin/python3
>> > >>>>>     
>> > >>>>> --target-list=x86_64-softmmu,mips-softmmu,mips64el-softmmu,aarch64-softmmu,arm-softmmu,s390x-softmmu,alpha-softmmu"
>> > >>>>>     > +        - CONFIG="--python=/usr/bin/python3
>> > >>>>>     
>> > >>>>> --target-list=x86_64-softmmu,mips-softmmu,mips64el-softmmu,aarch64-softmmu,arm-softmmu,s390x-softmmu,alpha-softmmu,riscv64-softmmu"
>> > >>>>>     >          - TEST_CMD="make check-acceptance"
>> > >>>>>     >        after_failure:
>> > >>>>>     >          - cat tests/results/latest/job.log
>> > >>>>>     > diff --git a/tests/acceptance/boot_linux_console.py
>> > >>>>>     b/tests/acceptance/boot_linux_console.py
>> > >>>>>     > index 3215950..0f638bc 100644
>> > >>>>>     > --- a/tests/acceptance/boot_linux_console.py
>> > >>>>>     > +++ b/tests/acceptance/boot_linux_console.py
>> > >>>>>     > @@ -354,3 +354,69 @@ class BootLinuxConsole(Test):
>> > >>>>>     >          self.vm.launch()
>> > >>>>>     >          console_pattern = 'Kernel command line: %s' %
>> > >>>>>     kernel_command_line
>> > >>>>>     >          self.wait_for_console_pattern(console_pattern)
>> > >>>>>     > +
>> > >>>>>     > +    def test_riscv64_virt(self):
>> > >>>>>     > +        """
>> > >>>>>     > +        :avocado: tags=arch:riscv64
>> > >>>>>     > +        :avocado: tags=machine:virt
>> > >>>>>     > +        """
>> > >>>>>     > +        deb_url = 
>> > >>>>> ('https://snapshot.debian.org/archive/debian/'
>> > >>>>>     > +                         
>> > >>>>> '20190424T171759Z/pool/main/b/binutils/'
>> > >>>>>     > +
>> > >>>>>      'binutils-riscv64-linux-gnu_2.32-8_amd64.deb')
>> > >>>>>     > +        deb_hash = 
>> > >>>>> ('7fe376fd4452696c03acd508d6d613ca553ea15e')
>> > >>>>>     > +        deb_path = self.fetch_asset(deb_url, 
>> > >>>>> asset_hash=deb_hash)
>> > >>>>>     > +        objcopy_path = '/usr/bin/riscv64-linux-gnu-objcopy'
>> > >>>>>     > +        objcopy_path = self.extract_from_deb(deb_path, 
>> > >>>>> objcopy_path)
>> > >>>>>     > +        libbfd_path =
>> > >>>>>     '/usr/lib/x86_64-linux-gnu/libbfd-2.32-riscv64.so
>> > >>>>>     <http://libbfd-2.32-riscv64.so>'
>> > >>>>>     > +        libbfd_path = self.extract_from_deb(deb_path, 
>> > >>>>> libbfd_path)
>> > >>>>>     > +        process.run('ls -al %s' % (objcopy_path))
>> > >>>>>     > +
>> > >>>>>     > +        deb_url = 
>> > >>>>> ('https://snapshot.debian.org/archive/debian/'
>> > >>>>>     > +                   '20190708T032337Z/pool/main/o/opensbi/'
>> > >>>>>     > +                   'opensbi_0.4-1_all.deb')
>> > >>>>>     > +        deb_hash = 
>> > >>>>> ('2319dcd702958291d323acf5649fd98a11d90112')
>> > >>>>>     > +        deb_path = self.fetch_asset(deb_url, 
>> > >>>>> asset_hash=deb_hash)
>> > >>>>>     > +        opensbi_path = ('/usr/lib/riscv64-linux-gnu/opensbi/'
>> > >>>>>     > +                        'qemu/virt/fw_jump.elf')
>> > >>>>>     > +        opensbi_path = self.extract_from_deb(deb_path, 
>> > >>>>> opensbi_path)
>> > >>>>>     > +
>> > >>>>>     > +        deb_url =
>> > >>>>>     ('https://snapshot.debian.org/archive/debian-ports/'
>> > >>>>>     > +                   
>> > >>>>> '20190620T095935Z/pool-riscv64/main/l/linux/'
>> > >>>>>     > +
>> > >>>>>      'linux-image-4.19.0-5-riscv64_4.19.37-4_riscv64.deb')
>> > >>>>>     > +        deb_hash = 
>> > >>>>> ('bf5b5680c41d92134d22caef4fbd277c5217e1f0')
>> > >>>>>     > +        deb_path = self.fetch_asset(deb_url, 
>> > >>>>> asset_hash=deb_hash)
>> > >>>>>     > +        kernel_path = '/boot/vmlinux-4.19.0-5-riscv64'
>> > >>>>>     > +        kernel_path = self.extract_from_deb(deb_path, 
>> > >>>>> kernel_path)
>> > >>>>>     > +        kimage_path = self.workdir + "/Image"
>> > >>>>>     > +        env = os.environ
>> > >>>>>     > +        env['LD_LIBRARY_PATH'] = ('%s:' %
>> > >>>>>     (os.path.dirname(libbfd_path)) +
>> > >>>>>     > +                                 env.get('LD_LIBRARY_PATH', 
>> > >>>>> ''))
>> > >>>>>     > +        process.run(('%s -O binary -O binary -R'
>> > >>>>>     > +                     '.note -R .note.gnu.build-id -R .comment 
>> > >>>>> -S
>> > >>>>>     %s %s') %
>> > >>>>>     > +                     (objcopy_path, kernel_path, kimage_path))
>> > >>>>>
>> > >>>>>     Please explain why you need to do that...
>> > >>>>>
>> > >>>>>     Also note these tests are run on other host architecture than 
>> > >>>>> x86, at
>> > >>>>>     least ppc64/s390x (so this won't work there).
>> > >>>>>
>> > >>>>> it is because riscv64 vmlinux doesn't have physical loading address
>> > >>>>> information and
>> > >>>>> depends on certain stage bootloader to move kernel raw data to 
>> > >>>>> specific
>> > >>>>> physical address (0x8020_0000)
>> > >>>>>
>> > >>>>> The vmlinux program headers are
>> > >>>>>   Type             Offset       VirtAddr                  PhysAddr
>> > >>>>>     FileSiz  MemSiz   Flg Align
>> > >>>>>   LOAD           0x001000 0xffffffe000000000 0x0000000000000000 
>> > >>>>> 0x0303a6
>> > >>>>> 0x0303a6 R E 0x1000
>> > >>>>>   LOAD           0x032000 0xffffffe000031000 0x0000000000031000 
>> > >>>>> 0x828f0c
>> > >>>>> 0x828f0c RWE 0x1000
>> > >>>>>   NOTE           0x85aec0 0xffffffe000859ec0 0x0000000000859ec0 
>> > >>>>> 0x00004c
>> > >>>>> 0x00004c R   0x4
>> > >>>>>
>> > >>>>> For legacy bootloader, bbl, It can handle payload in ELF format.
>> > >>>>> However, for the newer OpenSBI, it can only handle raw data and
>> > >>>>> debian doesn't provide linux image in raw data format.
>> > >>>
>> > >>> This doesn't sound right. You should just be able to say -kernel
>> > >>> <anything> and it should work. It has worked for me with vmlinux and
>> > >>> Image files when using master (not the 4.0 release but master/4.1).
>> > >>> Although all of my testing was with the 5.1 kernel, so maybe there is
>> > >>> a difference there?
>> > >>>
>> > >>> What isn't working for you? Can you include the errors and output of 
>> > >>> -d in_asm?
>> > >>>
>> > >>>
>> > >>> Alistair
>> > >>
>> > >>
>> > >> Hi Alistair,
>> > >>    I have come across error before starting target simulation.  What I 
>> > >> means is to execute
>> > >>             qemu-system-riscv64 -M virt -m 256M -nographic -bios 
>> > >> ./opensbi/build/platform/qemu/virt/firmware/fw_jump.elf -kernel vmlinux
>> > >>    then get the error message
>> > >>             rom: requested regions overlap (rom mrom.reset. 
>> > >> free=0x000000000001eb7c, addr=0x0000000000001000)
>> > >>             qemu-system-riscv64: rom check and register reset failed
>> > >
>> > > Hmmm.... I don't remember seeing this problem in my testing, but it
>> > > does seem to make sense.
>> > >
>> > > We have two options here:
>> > > 1. Use the Image file from Linux 5.0+ (my preferred option as 4.19 is
>> > > getting old)
>
>  I choose 4.19 version rather than 5.0+  because 4.19 is stable branch 
> maintained by Greg.
>  But it is ok to use 5.0 trunk, the latest version provided by debian.
>
>>
>> > > 2. Use the device loader (documented here:
>> > > https://github.com/riscv/opensbi/blob/master/docs/platform/qemu_virt.md)
>
>
> The document describes two user cases
>   1. fw_payload.elf
>       a. include the Image (raw data file, not elf)  into data section
>       b. -kernel load the each part to the correct address
>
>      the program header of  fw_payload.elf are
>      Type           Offset   VirtAddr           PhysAddr           FileSiz  
> MemSiz   Flg Align
>     LOAD           0x0000b0 0x0000000080000000 0x0000000080000000 0x00a008 
> 0x00b0b8 RWE 0x10               <-  fw_payload  main program
>     LOAD           0x00a0b8 0x0000000080200000 0x0000000080200000 0x8dbe20 
> 0x8dbe20 R E 0x1                    <-  arch/riscv/boot/Image   or u-boot.bin
>
>   2. fw_jump.elf
>       a. use  "-kernel <raw image> or -device  loader,file=<raw 
> Image>,addr=0x80200000" to load correct address
>       b. fw_jump.elf just jump to the entry address 0x80200000
>
>
>  The issues are
>    a. debian linux-image package only provides vmlinux.

This can be fixed by upgrading the kernel version. Fedora should have
a 5.2 kernel that you can use instead of Debian.

>        so I use objcopy in acceptance test
>    b. riscv64 vmlinux doesn't have correct PhysAddr and -kernel can't move 
> data to expected address
>        explicit load the Image to specific
>            -kernel raw_image   (if file is not elf or uImage,  
> riscv_kerner_loader treat it is as raw image and place it at 0x8020_0000)
>            -device loader,file=<raw_image>,0x80200000  (do the same thing)

Does the device loader not work?

Alistair

>
>   ELF should work only if it has correct loading information.
>
>   chihmin.chao
>>
>> >
>> > Nice doc, this example is exactly what I had in mind:
>> >
>> >   qemu-system-riscv64 -M virt -m 256M -nographic \
>> >         -kernel build/platform/qemu/virt/firmware/fw_jump.elf \
>> >         -device
>> > loader,file=<linux_build_directory>/arch/riscv/boot/Image,addr=0x80200000 \
>> >         -drive file=<path_to_linux_rootfs>,format=raw,id=hd0 \
>> >         -device virtio-blk-device,drive=hd0 \
>> >         -append "root=/dev/vda rw console=ttyS0"
>> >
>> > If the Image is an ELF, I assume we don't need to specify the address,
>> > because the device loader knows how to parse this format. Again,
>> > Alistair is the expert ;)
>>
>> Yep, an elf should just work.
>>
>  describe the issue above
>
>>
>> >
>> > If for some reason it only works with Linux 5.0+, let's use these!
>>
>> Only 5.0+ builds an elf.
>>
>> Alistair
>>
>> >
>> > Regards,
>> >
>> > Phil.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]