| Message ID | 20181012050757.6925-2-takahiro.akashi@linaro.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [1/2] efi_loader: rework fdt handling in distro boot script | expand |
Hi Takahiro, On Fri, 12 Oct 2018 14:07:57 +0900 AKASHI Takahiro <takahiro.akashi@linaro.org> wrote: > This variable, fdt_addr_t, is missing in the current qemu-arm.h while > it seems to be mandatory, at least, to run distro_bootcmd as expected. > So just add its definition. A size of 1MB would be enough. > In what way is this required for distro_bootcmd to work? At least in the past I've tested qemu_arm64_defconfig and EFI boot with the Fedora netinst image and it has worked fine in stock U-Boot. Note that these '-machine virt' based targets are slightly different from real hardware in the sense that instead of loading a .dtb file provided by the OS, the device tree is provided by QEMU. In the hunk below you can see "fdt_addr=0x40000000\0" providing the address of the QEMU-provided device tree which the distro scripts should be using. I guess in principle having ${fdt_addr_r} set as well shouldn't hurt and might be used for testing/unusual purposes. Glancing at cmd/pxe.c, there is a problem though, in that if ${fdt_addr_r} were defined, a PXE file using the FDTDIR directive would attempt loading a dtb file named "<NULL>-qemu-arm.dtb" instead of falling back to using ${fdt_addr}. That bug would need to be fixed first before applying this patch. > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> > --- > include/configs/qemu-arm.h | 1 + > 1 file changed, 1 insertion(+) > > diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h > index 91fb8d47edf8..0e66f946dde5 100644 > --- a/include/configs/qemu-arm.h > +++ b/include/configs/qemu-arm.h > @@ -55,6 +55,7 @@ > "fdt_high=0xffffffff\0" \ > "initrd_high=0xffffffff\0" \ > "fdt_addr=0x40000000\0" \ > + "fdt_addr_r=0x40100000\0" \ > "scriptaddr=0x40200000\0" \ > "pxefile_addr_r=0x40300000\0" \ > "kernel_addr_r=0x40400000\0" \
On Mon, Oct 15, 2018 at 04:01:06AM +0300, Tuomas Tynkkynen wrote: > Hi Takahiro, > > On Fri, 12 Oct 2018 14:07:57 +0900 > AKASHI Takahiro <takahiro.akashi@linaro.org> wrote: > > > This variable, fdt_addr_t, is missing in the current qemu-arm.h while > > it seems to be mandatory, at least, to run distro_bootcmd as expected. > > So just add its definition. A size of 1MB would be enough. > > > > In what way is this required for distro_bootcmd to work? At least in the > past I've tested qemu_arm64_defconfig and EFI boot with the Fedora > netinst image and it has worked fine in stock U-Boot. > > Note that these '-machine virt' based targets are slightly different > from real hardware in the sense that instead of loading a .dtb file > provided by the OS, the device tree is provided by QEMU. In the hunk > below you can see "fdt_addr=0x40000000\0" providing the address of > the QEMU-provided device tree which the distro scripts should be > using. Yep, I know that. I was not clear; what distro bootcmd, or more specifically scan_dev_for_efi, tries to do regarding fdt handling is that, if "fdtfile" variable is defined, it supersedes qemu's own dtb (that is what "fdt_addr" points to), but this trick doesn't work expectedly if "fdt_addr_r" variable is not defined. > I guess in principle having ${fdt_addr_r} set as well shouldn't hurt and > might be used for testing/unusual purposes. I don't know whether the case is "unsual" or not. But doc/README.distro cleary says that fdt_addr_r is *mandatory*. If u-boot works without it, it's a bug, otherwise we must correct the doc (or scan_dev_for_efi in the first place?). Thanks, -Takahiro Akashi > Glancing at cmd/pxe.c, > there is a problem though, in that if ${fdt_addr_r} were defined, a PXE > file using the FDTDIR directive would attempt loading a dtb file named > "<NULL>-qemu-arm.dtb" instead of falling back to using ${fdt_addr}. > That bug would need to be fixed first before applying this patch. > > > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> > > --- > > include/configs/qemu-arm.h | 1 + > > 1 file changed, 1 insertion(+) > > > > diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h > > index 91fb8d47edf8..0e66f946dde5 100644 > > --- a/include/configs/qemu-arm.h > > +++ b/include/configs/qemu-arm.h > > @@ -55,6 +55,7 @@ > > "fdt_high=0xffffffff\0" \ > > "initrd_high=0xffffffff\0" \ > > "fdt_addr=0x40000000\0" \ > > + "fdt_addr_r=0x40100000\0" \ > > "scriptaddr=0x40200000\0" \ > > "pxefile_addr_r=0x40300000\0" \ > > "kernel_addr_r=0x40400000\0" \ >
On 15.10.18 07:14, AKASHI Takahiro wrote: > On Mon, Oct 15, 2018 at 04:01:06AM +0300, Tuomas Tynkkynen wrote: >> Hi Takahiro, >> >> On Fri, 12 Oct 2018 14:07:57 +0900 >> AKASHI Takahiro <takahiro.akashi@linaro.org> wrote: >> >>> This variable, fdt_addr_t, is missing in the current qemu-arm.h while >>> it seems to be mandatory, at least, to run distro_bootcmd as expected. >>> So just add its definition. A size of 1MB would be enough. >>> >> >> In what way is this required for distro_bootcmd to work? At least in the >> past I've tested qemu_arm64_defconfig and EFI boot with the Fedora >> netinst image and it has worked fine in stock U-Boot. >> >> Note that these '-machine virt' based targets are slightly different >> from real hardware in the sense that instead of loading a .dtb file >> provided by the OS, the device tree is provided by QEMU. In the hunk >> below you can see "fdt_addr=0x40000000\0" providing the address of >> the QEMU-provided device tree which the distro scripts should be >> using. > > Yep, I know that. > I was not clear; what distro bootcmd, or more specifically scan_dev_for_efi, > tries to do regarding fdt handling is that, if "fdtfile" variable is defined, > it supersedes qemu's own dtb (that is what "fdt_addr" points to), but > this trick doesn't work expectedly if "fdt_addr_r" variable is not defined. > >> I guess in principle having ${fdt_addr_r} set as well shouldn't hurt and >> might be used for testing/unusual purposes. > > I don't know whether the case is "unsual" or not. > But doc/README.distro cleary says that fdt_addr_r is *mandatory*. > If u-boot works without it, it's a bug, otherwise we must correct > the doc (or scan_dev_for_efi in the first place?). I agree. Boards that support distro boot *must* provide fdt_addr_r. Otherwise distro scripts may fail unexpectedly because they assume the variable is present. > > Thanks, > -Takahiro Akashi > >> Glancing at cmd/pxe.c, >> there is a problem though, in that if ${fdt_addr_r} were defined, a PXE >> file using the FDTDIR directive would attempt loading a dtb file named >> "<NULL>-qemu-arm.dtb" instead of falling back to using ${fdt_addr}. >> That bug would need to be fixed first before applying this patch. Well, and that load will fail and everyone's happy, no? IMHO we should fall back to $fdtcontroladdr always, but the pxe boot path is not something I would endorse onto anyone (what you want as PXE really is called 'dhcp' in the efi_loader distro boot world) Alex
Hi Alexander, On Tue, 16 Oct 2018 15:04:26 +0200 Alexander Graf <agraf@suse.de> wrote: ... > > > >> Glancing at cmd/pxe.c, > >> there is a problem though, in that if ${fdt_addr_r} were defined, > >> a PXE file using the FDTDIR directive would attempt loading a dtb > >> file named "<NULL>-qemu-arm.dtb" instead of falling back to using > >> ${fdt_addr}. That bug would need to be fixed first before applying > >> this patch. > > Well, and that load will fail and everyone's happy, no? No, because currently if DTB loading from the filesystem/tftp is attempted and it fails, it aborts the boot. It doesn't matter if it's via a 'FDT' or 'FDTDIR' directive. In the case of typical hardware that's probably the desired behaviour. I guess the qemu_arm + FDTDIR case can be fixed by not attempting a .dtb load if the default fdtfile is not known due to $soc or $board being unset. At least I doubt that some other board could be relying on a filename containing literally "<NULL>" :) > IMHO we should > fall back to $fdtcontroladdr always FWIW, to me the idea of passing $fdtcontroladdr to the OS has always filled me with dread. On top of the usual backwards- and forwards-compatibility problems that happen when mixing and matching kernels and DTBs from different releases, you now have to deal with issues like U-Boot specific .dts that are majorly diverged from Linux ones, or where the .dts is otherwise from Linux but the U-Boot specific additions break it for Linux, or where the .dts used is wrong for the specific hardware revision but close enough for U-Boot's purposes, and so on... - Tuomas
On 18.10.18 00:25, Tuomas Tynkkynen wrote: > Hi Alexander, > > On Tue, 16 Oct 2018 15:04:26 +0200 > Alexander Graf <agraf@suse.de> wrote: > > ... >>> >>>> Glancing at cmd/pxe.c, >>>> there is a problem though, in that if ${fdt_addr_r} were defined, >>>> a PXE file using the FDTDIR directive would attempt loading a dtb >>>> file named "<NULL>-qemu-arm.dtb" instead of falling back to using >>>> ${fdt_addr}. That bug would need to be fixed first before applying >>>> this patch. >> >> Well, and that load will fail and everyone's happy, no? > > No, because currently if DTB loading from the filesystem/tftp is > attempted and it fails, it aborts the boot. It doesn't matter if it's > via a 'FDT' or 'FDTDIR' directive. In the case of typical hardware > that's probably the desired behaviour. > > I guess the qemu_arm + FDTDIR case can be fixed by not attempting > a .dtb load if the default fdtfile is not known due to $soc or $board > being unset. At least I doubt that some other board could be relying > on a filename containing literally "<NULL>" :) Well - IIRC $soc and $board should also always be defined ;). So yet another thing to fix in the QEMU port. > >> IMHO we should >> fall back to $fdtcontroladdr always > > FWIW, to me the idea of passing $fdtcontroladdr to the OS has always > filled me with dread. On top of the usual backwards- and > forwards-compatibility problems that happen when mixing and matching > kernels and DTBs from different releases, you now have to deal with That's something that we seriously as a community need to get sorted out. We're pushing hard for it in the EBBR context. The fact that people are afraid of putting *hardware desciption* into their firmware is just mind boggling. > issues like U-Boot specific .dts that are majorly diverged from Linux > ones, or where the .dts is otherwise from Linux but the U-Boot specific These case should really be the minority. And if you see those, please fix them. > additions break it for Linux, or where the .dts used is wrong for the I've never seen a case where a U-Boot addition broke the DT for Linux. > specific hardware revision but close enough for U-Boot's purposes, > and so on... Again, something that just needs fixing. Device trees belong to the entity that knows hardware, not to the OS. Otherwise you lose the abstraction layer that DT gives you and you lose the ability to run "generic" kernels. And of course you break the ecosystem, because now good luck running BSD, your own little toy OS, etc ;) Alex
On Thu, Oct 18, 2018 at 09:25:04AM +0200, Alexander Graf wrote: > > > On 18.10.18 00:25, Tuomas Tynkkynen wrote: > > Hi Alexander, > > > > On Tue, 16 Oct 2018 15:04:26 +0200 > > Alexander Graf <agraf@suse.de> wrote: > > > > ... > >>> > >>>> Glancing at cmd/pxe.c, > >>>> there is a problem though, in that if ${fdt_addr_r} were defined, > >>>> a PXE file using the FDTDIR directive would attempt loading a dtb > >>>> file named "<NULL>-qemu-arm.dtb" instead of falling back to using > >>>> ${fdt_addr}. That bug would need to be fixed first before applying > >>>> this patch. > >> > >> Well, and that load will fail and everyone's happy, no? > > > > No, because currently if DTB loading from the filesystem/tftp is > > attempted and it fails, it aborts the boot. It doesn't matter if it's > > via a 'FDT' or 'FDTDIR' directive. In the case of typical hardware > > that's probably the desired behaviour. > > > > I guess the qemu_arm + FDTDIR case can be fixed by not attempting > > a .dtb load if the default fdtfile is not known due to $soc or $board > > being unset. At least I doubt that some other board could be relying > > on a filename containing literally "<NULL>" :) > > Well - IIRC $soc and $board should also always be defined ;). So yet > another thing to fix in the QEMU port. See my patch below. Are you happy with it? (qemu-qemu-arm.dtb doesn't make sense to me, though :) > > > >> IMHO we should > >> fall back to $fdtcontroladdr always > > > > FWIW, to me the idea of passing $fdtcontroladdr to the OS has always > > filled me with dread. On top of the usual backwards- and > > forwards-compatibility problems that happen when mixing and matching > > kernels and DTBs from different releases, you now have to deal with > > That's something that we seriously as a community need to get sorted > out. We're pushing hard for it in the EBBR context. The fact that people > are afraid of putting *hardware desciption* into their firmware is just > mind boggling. I don't have a strong opinion, but it seems to me that 'fall-back' approach is quite normal. If it doesn't work on a specific board, 'fdt_addr' should be defined. Thanks, -Takahiro Akashi > > issues like U-Boot specific .dts that are majorly diverged from Linux > > ones, or where the .dts is otherwise from Linux but the U-Boot specific > > These case should really be the minority. And if you see those, please > fix them. > > > additions break it for Linux, or where the .dts used is wrong for the > > I've never seen a case where a U-Boot addition broke the DT for Linux. > > > specific hardware revision but close enough for U-Boot's purposes, > > and so on... > > Again, something that just needs fixing. Device trees belong to the > entity that knows hardware, not to the OS. Otherwise you lose the > abstraction layer that DT gives you and you lose the ability to run > "generic" kernels. And of course you break the ecosystem, because now > good luck running BSD, your own little toy OS, etc ;) > > > Alex ===8<=== From 47b26a86359a3b612e890f2ca2d5d20092f9f4e1 Mon Sep 17 00:00:00 2001 From: AKASHI Takahiro <takahiro.akashi@linaro.org> Date: Fri, 19 Oct 2018 15:22:02 +0900 Subject: [PATCH] arm: qemu: rework Kconfig Define SYS_SOC and move SYS_* to a canonical place (under board). Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> --- arch/arm/mach-qemu/Kconfig | 18 ++++++++++-------- board/emulation/Kconfig | 2 ++ board/emulation/qemu-arm/Kconfig | 9 +++++++++ 3 files changed, 21 insertions(+), 8 deletions(-) create mode 100644 board/emulation/qemu-arm/Kconfig diff --git a/arch/arm/mach-qemu/Kconfig b/arch/arm/mach-qemu/Kconfig index a2e4b98b8887..d75a95183a75 100644 --- a/arch/arm/mach-qemu/Kconfig +++ b/arch/arm/mach-qemu/Kconfig @@ -3,22 +3,24 @@ if ARCH_QEMU config SYS_VENDOR default "emulation" -config SYS_BOARD - default "qemu-arm" +config SYS_SOC + default "qemu" -config SYS_CONFIG_NAME - default "qemu-arm" - -endif +choice + prompt "QEMU cpu type" config TARGET_QEMU_ARM_32BIT - bool "Support qemu_arm" + bool "Arm" depends on ARCH_QEMU select ARCH_SUPPORT_PSCI select CPU_V7A select SYS_ARCH_TIMER config TARGET_QEMU_ARM_64BIT - bool "Support qemu_arm64" + bool "AArch64" depends on ARCH_QEMU select ARM64 + +endchoice + +endif diff --git a/board/emulation/Kconfig b/board/emulation/Kconfig index f821458fa6ac..597e926cdb11 100644 --- a/board/emulation/Kconfig +++ b/board/emulation/Kconfig @@ -27,4 +27,6 @@ endchoice source "board/emulation/qemu-x86/Kconfig" +source "board/emulation/qemu-arm/Kconfig" + endif diff --git a/board/emulation/qemu-arm/Kconfig b/board/emulation/qemu-arm/Kconfig new file mode 100644 index 000000000000..db8b2a4dfae2 --- /dev/null +++ b/board/emulation/qemu-arm/Kconfig @@ -0,0 +1,9 @@ +if TARGET_QEMU_ARM_32BIT || TARGET_QEMU_ARM_64BIT + +config SYS_BOARD + default "qemu-arm" + +config SYS_CONFIG_NAME + default "qemu-arm" + +endif
On 19.10.18 08:33, AKASHI Takahiro wrote: > On Thu, Oct 18, 2018 at 09:25:04AM +0200, Alexander Graf wrote: >> >> >> On 18.10.18 00:25, Tuomas Tynkkynen wrote: >>> Hi Alexander, >>> >>> On Tue, 16 Oct 2018 15:04:26 +0200 >>> Alexander Graf <agraf@suse.de> wrote: >>> >>> ... >>>>> >>>>>> Glancing at cmd/pxe.c, >>>>>> there is a problem though, in that if ${fdt_addr_r} were defined, >>>>>> a PXE file using the FDTDIR directive would attempt loading a dtb >>>>>> file named "<NULL>-qemu-arm.dtb" instead of falling back to using >>>>>> ${fdt_addr}. That bug would need to be fixed first before applying >>>>>> this patch. >>>> >>>> Well, and that load will fail and everyone's happy, no? >>> >>> No, because currently if DTB loading from the filesystem/tftp is >>> attempted and it fails, it aborts the boot. It doesn't matter if it's >>> via a 'FDT' or 'FDTDIR' directive. In the case of typical hardware >>> that's probably the desired behaviour. >>> >>> I guess the qemu_arm + FDTDIR case can be fixed by not attempting >>> a .dtb load if the default fdtfile is not known due to $soc or $board >>> being unset. At least I doubt that some other board could be relying >>> on a filename containing literally "<NULL>" :) >> >> Well - IIRC $soc and $board should also always be defined ;). So yet >> another thing to fix in the QEMU port. > > See my patch below. Are you happy with it? > (qemu-qemu-arm.dtb doesn't make sense to me, though :) > >>> >>>> IMHO we should >>>> fall back to $fdtcontroladdr always >>> >>> FWIW, to me the idea of passing $fdtcontroladdr to the OS has always >>> filled me with dread. On top of the usual backwards- and >>> forwards-compatibility problems that happen when mixing and matching >>> kernels and DTBs from different releases, you now have to deal with >> >> That's something that we seriously as a community need to get sorted >> out. We're pushing hard for it in the EBBR context. The fact that people >> are afraid of putting *hardware desciption* into their firmware is just >> mind boggling. > > I don't have a strong opinion, but it seems to me that 'fall-back' approach > is quite normal. If it doesn't work on a specific board, 'fdt_addr' > should be defined. > > Thanks, > -Takahiro Akashi > > >>> issues like U-Boot specific .dts that are majorly diverged from Linux >>> ones, or where the .dts is otherwise from Linux but the U-Boot specific >> >> These case should really be the minority. And if you see those, please >> fix them. >> >>> additions break it for Linux, or where the .dts used is wrong for the >> >> I've never seen a case where a U-Boot addition broke the DT for Linux. >> >>> specific hardware revision but close enough for U-Boot's purposes, >>> and so on... >> >> Again, something that just needs fixing. Device trees belong to the >> entity that knows hardware, not to the OS. Otherwise you lose the >> abstraction layer that DT gives you and you lose the ability to run >> "generic" kernels. And of course you break the ecosystem, because now >> good luck running BSD, your own little toy OS, etc ;) >> >> >> Alex > > ===8<=== > From 47b26a86359a3b612e890f2ca2d5d20092f9f4e1 Mon Sep 17 00:00:00 2001 > From: AKASHI Takahiro <takahiro.akashi@linaro.org> > Date: Fri, 19 Oct 2018 15:22:02 +0900 > Subject: [PATCH] arm: qemu: rework Kconfig > > Define SYS_SOC and move SYS_* to a canonical place (under board). > > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> Looks good enough to me, but it's up to Tuomas to double check and take as he's the qemu-arm maintainer :). It also usually helps to post the patch as a new message, not buried inside a thread ;). Alex
On Fri, Oct 19, 2018 at 09:46:51AM +0200, Alexander Graf wrote: > > > On 19.10.18 08:33, AKASHI Takahiro wrote: > > On Thu, Oct 18, 2018 at 09:25:04AM +0200, Alexander Graf wrote: > >> > >> > >> On 18.10.18 00:25, Tuomas Tynkkynen wrote: > >>> Hi Alexander, > >>> > >>> On Tue, 16 Oct 2018 15:04:26 +0200 > >>> Alexander Graf <agraf@suse.de> wrote: > >>> > >>> ... > >>>>> > >>>>>> Glancing at cmd/pxe.c, > >>>>>> there is a problem though, in that if ${fdt_addr_r} were defined, > >>>>>> a PXE file using the FDTDIR directive would attempt loading a dtb > >>>>>> file named "<NULL>-qemu-arm.dtb" instead of falling back to using > >>>>>> ${fdt_addr}. That bug would need to be fixed first before applying > >>>>>> this patch. > >>>> > >>>> Well, and that load will fail and everyone's happy, no? > >>> > >>> No, because currently if DTB loading from the filesystem/tftp is > >>> attempted and it fails, it aborts the boot. It doesn't matter if it's > >>> via a 'FDT' or 'FDTDIR' directive. In the case of typical hardware > >>> that's probably the desired behaviour. > >>> > >>> I guess the qemu_arm + FDTDIR case can be fixed by not attempting > >>> a .dtb load if the default fdtfile is not known due to $soc or $board > >>> being unset. At least I doubt that some other board could be relying > >>> on a filename containing literally "<NULL>" :) > >> > >> Well - IIRC $soc and $board should also always be defined ;). So yet > >> another thing to fix in the QEMU port. > > > > See my patch below. Are you happy with it? > > (qemu-qemu-arm.dtb doesn't make sense to me, though :) > > > >>> > >>>> IMHO we should > >>>> fall back to $fdtcontroladdr always > >>> > >>> FWIW, to me the idea of passing $fdtcontroladdr to the OS has always > >>> filled me with dread. On top of the usual backwards- and > >>> forwards-compatibility problems that happen when mixing and matching > >>> kernels and DTBs from different releases, you now have to deal with > >> > >> That's something that we seriously as a community need to get sorted > >> out. We're pushing hard for it in the EBBR context. The fact that people > >> are afraid of putting *hardware desciption* into their firmware is just > >> mind boggling. > > > > I don't have a strong opinion, but it seems to me that 'fall-back' approach > > is quite normal. If it doesn't work on a specific board, 'fdt_addr' > > should be defined. > > > > Thanks, > > -Takahiro Akashi > > > > > >>> issues like U-Boot specific .dts that are majorly diverged from Linux > >>> ones, or where the .dts is otherwise from Linux but the U-Boot specific > >> > >> These case should really be the minority. And if you see those, please > >> fix them. > >> > >>> additions break it for Linux, or where the .dts used is wrong for the > >> > >> I've never seen a case where a U-Boot addition broke the DT for Linux. > >> > >>> specific hardware revision but close enough for U-Boot's purposes, > >>> and so on... > >> > >> Again, something that just needs fixing. Device trees belong to the > >> entity that knows hardware, not to the OS. Otherwise you lose the > >> abstraction layer that DT gives you and you lose the ability to run > >> "generic" kernels. And of course you break the ecosystem, because now > >> good luck running BSD, your own little toy OS, etc ;) > >> > >> > >> Alex > > > > ===8<=== > > From 47b26a86359a3b612e890f2ca2d5d20092f9f4e1 Mon Sep 17 00:00:00 2001 > > From: AKASHI Takahiro <takahiro.akashi@linaro.org> > > Date: Fri, 19 Oct 2018 15:22:02 +0900 > > Subject: [PATCH] arm: qemu: rework Kconfig > > > > Define SYS_SOC and move SYS_* to a canonical place (under board). > > > > Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> > > Looks good enough to me, but it's up to Tuomas to double check and take > as he's the qemu-arm maintainer :). > > It also usually helps to post the patch as a new message, not buried > inside a thread ;). OK. Since I found a small bug, I will send out a new patch separately. -Takahiro Akashi > > Alex
On Thu, 18 Oct 2018 09:25:04 +0200 Alexander Graf <agraf@suse.de> wrote: > On 18.10.18 00:25, Tuomas Tynkkynen wrote: > > Hi Alexander, > > > > On Tue, 16 Oct 2018 15:04:26 +0200 > > Alexander Graf <agraf@suse.de> wrote: > > > > ... > >>> > >>>> Glancing at cmd/pxe.c, > >>>> there is a problem though, in that if ${fdt_addr_r} were defined, > >>>> a PXE file using the FDTDIR directive would attempt loading a dtb > >>>> file named "<NULL>-qemu-arm.dtb" instead of falling back to using > >>>> ${fdt_addr}. That bug would need to be fixed first before > >>>> applying this patch. > >> > >> Well, and that load will fail and everyone's happy, no? > > > > No, because currently if DTB loading from the filesystem/tftp is > > attempted and it fails, it aborts the boot. It doesn't matter if > > it's via a 'FDT' or 'FDTDIR' directive. In the case of typical > > hardware that's probably the desired behaviour. > > > > I guess the qemu_arm + FDTDIR case can be fixed by not attempting > > a .dtb load if the default fdtfile is not known due to $soc or > > $board being unset. At least I doubt that some other board could be > > relying on a filename containing literally "<NULL>" :) > > Well - IIRC $soc and $board should also always be defined ;). So yet > another thing to fix in the QEMU port. As far as I know $soc is only used for computing the default $fdtfile, but we explicitly don't want to compute one for qemu_arm. > > > >> IMHO we should > >> fall back to $fdtcontroladdr always > > > > FWIW, to me the idea of passing $fdtcontroladdr to the OS has always > > filled me with dread. On top of the usual backwards- and > > forwards-compatibility problems that happen when mixing and matching > > kernels and DTBs from different releases, you now have to deal > > with > > That's something that we seriously as a community need to get sorted > out. We're pushing hard for it in the EBBR context. The fact that > people are afraid of putting *hardware desciption* into their > firmware is just mind boggling. But until/if that is solved anyone trying to rely on $fdtcontroladdr will get burned, or at least that's what the experience for sunxi sounds like where stuff like adding new regulators to DT breaks old kernels that lack the regulator driver. (And where the proposed solution to that was some hack that just uses fixed regulators instead, throwing the hardware description aspect out the window). And it doesn't look like an isolated situation, you'd have got equally hosed e.g when RPi switched to using the sdhost IP in place of the other MMC controller, or when Tegra landed USB3 driver support thus switched the .dts to using the XUSB controllers over the USB2 ones. > > issues like U-Boot specific .dts that are majorly diverged from > > Linux ones, or where the .dts is otherwise from Linux but the > > U-Boot specific > > These case should really be the minority. And if you see those, please > fix them. Well, for the case of Tegra124, being able to use a recent unmodified .dts from Linux for U-Boot would require writing an USB3 driver in order to not break USB functionality. That's not exactly a simple fix. > > additions break it for Linux, or where the .dts used is wrong for > > the > > I've never seen a case where a U-Boot addition broke the DT for Linux. > See e.g. 96a82d33f8c5bd3e90098b540349b7b9e43e27da fixing such instances. > > specific hardware revision but close enough for U-Boot's purposes, > > and so on... > > Again, something that just needs fixing. Device trees belong to the > entity that knows hardware, not to the OS. Otherwise you lose the > abstraction layer that DT gives you and you lose the ability to run > "generic" kernels. And of course you break the ecosystem, because now > good luck running BSD, your own little toy OS, etc ;) > It's been possible to make generic images using the U-Boot distro stuff (where U-Boot knows just the .dtb filename to load from the ones installed by kernel's `make dtbs_install`) for over three years now. Sure it doesn't fulfill the "DT doesn't belong to the OS" goal... but given that features (incl. DT bindings) get developed in vendor kernels first but upstream won't typically accept then unchanged, having a single .dtb that satisfies both has been a logical impossibility so far. - Tuomas
diff --git a/include/configs/qemu-arm.h b/include/configs/qemu-arm.h index 91fb8d47edf8..0e66f946dde5 100644 --- a/include/configs/qemu-arm.h +++ b/include/configs/qemu-arm.h @@ -55,6 +55,7 @@ "fdt_high=0xffffffff\0" \ "initrd_high=0xffffffff\0" \ "fdt_addr=0x40000000\0" \ + "fdt_addr_r=0x40100000\0" \ "scriptaddr=0x40200000\0" \ "pxefile_addr_r=0x40300000\0" \ "kernel_addr_r=0x40400000\0" \
This variable, fdt_addr_t, is missing in the current qemu-arm.h while it seems to be mandatory, at least, to run distro_bootcmd as expected. So just add its definition. A size of 1MB would be enough. Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org> --- include/configs/qemu-arm.h | 1 + 1 file changed, 1 insertion(+)