Re: [U-Boot] [PATCH] tegra2: move tegra2 SoC code to arch/arm/cpu/tegra2-common
Hi Allen, On Thu, Apr 19, 2012 at 12:22 PM, Allen Martin wrote: > On Thu, Apr 19, 2012 at 11:27:05AM -0700, Simon Glass wrote: > > > > > > This is part of an upcoming patch set to move all armv4t code out of > > > tegra2 u-boot and put it into an SPL, but I wanted to get some early > > > feedback on this patch. This moves most of the tegra2 SoC code from > > > arch/arm/cpu/armv7 to a new directory arch/arm/cpu/tegra2-common. > > > This code will be shared between the armv7 and armv4t builds of > > > tegra2. > > > > I am not sure whether moving things up a level is OK. It makes some > > sense to me but is different from what other boards do. > > Agreed I couldn't find any other examples, but AFAIK no other board is > trying to do what we do (sharing code between 2 different CPU > models). The other options I contemplated were: > > -Have the armv4t build reach up and over into armv7 > -Move the code out to board/nvidia > > Both of these seemed worse > I've thought about this a bit more. To me you have a bit of a unique problem in that you need to build the code as both ARMv4T and ARMv7. The way it works at the moment is pretty simple - we just mark a few files that must be build with ARMv4T and ARMv7 is happy with that also. I believe the point of the 'cpu' directory is to separate out common code between different architectures. It seems you want a chip which uses two different architectures. That might be unique in the U-Boot world - I note that other SOCs with different chips use ARMv7 for all of them. But really, who cares what architecture you actually use? The code clearly doesn't include ARMv7-isms otherwise it wouldn't build with current compile options. It does work... So my suggestion is that you just continue as now, and build the relevant code for ARMv4T. I don't see a compelling reason to move it, least of all into a no-man's land at the same level as 'cpu'. > > > Secondly, you will at some point add Tegra 3 support I suppose. In > > that case much of this code will become common anyway. Should you do > > that split (tegra-common, tegra2-common) now or later? > > I'd like to keep the tegra3 changes separate since that will involve > refactoring of much of this code. For the SPL build it's mostly > hidden in the config and Makefiles. > > I thought about doing exactly what you said to reduce the amount of > code I needed to pull into the SPL build, but what I found was that > because the SPL needs to initialize some regulators, clocks, and > pinmux to bring up the A9 and UART it ends up needing much of the > tegra2 SoC code, so there's not a lot of opportunity to minize the > code set for the SPL build beyond the CONFIG options to disable > drivers that area already there. > OK. Please note that we do have a refactor in the Chromium tree to move things into tegra-common and add T30 support. We hope to get this upstream once the T20 stuff is there (NAND, keyboard and LCD are still pending). Regards, Simon > > -Allen > nvpublic > ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH] tegra2: move tegra2 SoC code to arch/arm/cpu/tegra2-common
On Mon, Apr 30, 2012 at 11:31:44AM -0700, Simon Glass wrote: > > -Have the armv4t build reach up and over into armv7 > -Move the code out to board/nvidia > > Both of these seemed worse > > I've thought about this a bit more. To me you have a bit of a unique problem > in that you need to build the code as both ARMv4T and ARMv7. The way it works > at the moment is pretty simple - we just mark a few files that must be build > with ARMv4T and ARMv7 is happy with that also. > > I believe the point of the 'cpu' directory is to separate out common code > between different architectures. It seems you want a chip which uses two > different architectures. That might be unique in the U-Boot world - I note > that other SOCs with different chips use ARMv7 for all of them. > > But really, who cares what architecture you actually use? The code clearly > doesn't include ARMv7-isms otherwise it wouldn't build with current compile > options. It does work... > > So my suggestion is that you just continue as now, and build the relevant > code for ARMv4T. I don't see a compelling reason to move it, least of all > into a no-man's land at the same level as 'cpu'. > The problem that pushed me over the edge on this was the patch series to add thumb support for ARM. The linker has to insert interworking code for the cross ARM/thumb calls. If it is linking armv4t and armv7 objects together it will promote them all to armv7 style interworking which uses instructions that are illegal on armv4t. The only fix for this is to link armv4t objects separately or not support thumb, and we would really like to support thumb. Also it's not true that the code in arch/arm/cpu/armv7 doesn't include any armv7'isms, we're just #ifdef around them for the tegra and call them later in the boot sequence when we know we're running on the A9. We also have to code the init sequence extremely carefully to make sure we don't call any armv7 code on the armv4t. It's extremely fragile as any changes to start.S or the ordering of the init sequence can (and have) broken us. -Allen -- nvpublic ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH] tegra2: move tegra2 SoC code to arch/arm/cpu/tegra2-common
Hi Allen, On Mon, Apr 30, 2012 at 12:52 PM, Allen Martin wrote: > On Mon, Apr 30, 2012 at 11:31:44AM -0700, Simon Glass wrote: > > > > -Have the armv4t build reach up and over into armv7 > > -Move the code out to board/nvidia > > > > Both of these seemed worse > > > > I've thought about this a bit more. To me you have a bit of a unique > problem in that you need to build the code as both ARMv4T and ARMv7. The > way it works at the moment is pretty simple - we just mark a few files that > must be build with ARMv4T and ARMv7 is happy with that also. > > > > I believe the point of the 'cpu' directory is to separate out common > code between different architectures. It seems you want a chip which uses > two different architectures. That might be unique in the U-Boot world - I > note that other SOCs with different chips use ARMv7 for all of them. > > > > But really, who cares what architecture you actually use? The code > clearly doesn't include ARMv7-isms otherwise it wouldn't build with current > compile options. It does work... > > > > So my suggestion is that you just continue as now, and build the > relevant code for ARMv4T. I don't see a compelling reason to move it, least > of all into a no-man's land at the same level as 'cpu'. > > > > The problem that pushed me over the edge on this was the patch series > to add thumb support for ARM. The linker has to insert interworking > code for the cross ARM/thumb calls. If it is linking armv4t and armv7 > objects together it will promote them all to armv7 style interworking > which uses instructions that are illegal on armv4t. The only fix for > this is to link armv4t objects separately or not support thumb, and we > would really like to support thumb. > OK I just remembered this thread... The Thumb reason makes good sense to me. That said, you could arrange to avoid interworked calls and have all the ARMv4 code built with the same flags (and not allow the ARM7TDMI to enter board_init_f()). So perhaps another way of putting it is that you are using SPL as a convenient and existing way to allow you to built two parts of U-Boot with separate flags. > > Also it's not true that the code in arch/arm/cpu/armv7 doesn't include > any armv7'isms, we're just #ifdef around them for the tegra and call > them later in the boot sequence when we know we're running on the A9. > We also have to code the init sequence extremely carefully to make > sure we don't call any armv7 code on the armv4t. It's extremely > fragile as any changes to start.S or the ordering of the init sequence > can (and have) broken us. > See my other mail about this. I think you overstate things a little :-) But anyway, I think I understand the desirability for this now. > > -Allen > -- > nvpublic > Regards, Simon ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH] tegra2: move tegra2 SoC code to arch/arm/cpu/tegra2-common
+Albert Hi Allen, On Wed, Apr 18, 2012 at 3:46 PM, Allen Martin wrote: > In preparation for splitting out the armv4t code from tegra2, move the > tegra2 SoC code to arch/arm/cpu/tegra2-common. This code will be > compiled armv4t for the arm7tdmi and armv7 for the cortex A9. > > Signed-off-by: Allen Martin > --- > Resending this with "git format-patch -M" to make it more readable > and removing cover letter > > This is part of an upcoming patch set to move all armv4t code out of > tegra2 u-boot and put it into an SPL, but I wanted to get some early > feedback on this patch. This moves most of the tegra2 SoC code from > arch/arm/cpu/armv7 to a new directory arch/arm/cpu/tegra2-common. > This code will be shared between the armv7 and armv4t builds of > tegra2. I am not sure whether moving things up a level is OK. It makes some sense to me but is different from what other boards do. Secondly, you will at some point add Tegra 3 support I suppose. In that case much of this code will become common anyway. Should you do that split (tegra-common, tegra2-common) now or later? Regards, Simon ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Re: [U-Boot] [PATCH] tegra2: move tegra2 SoC code to arch/arm/cpu/tegra2-common
On Thu, Apr 19, 2012 at 11:27:05AM -0700, Simon Glass wrote: > > > > This is part of an upcoming patch set to move all armv4t code out of > > tegra2 u-boot and put it into an SPL, but I wanted to get some early > > feedback on this patch. This moves most of the tegra2 SoC code from > > arch/arm/cpu/armv7 to a new directory arch/arm/cpu/tegra2-common. > > This code will be shared between the armv7 and armv4t builds of > > tegra2. > > I am not sure whether moving things up a level is OK. It makes some > sense to me but is different from what other boards do. Agreed I couldn't find any other examples, but AFAIK no other board is trying to do what we do (sharing code between 2 different CPU models). The other options I contemplated were: -Have the armv4t build reach up and over into armv7 -Move the code out to board/nvidia Both of these seemed worse > Secondly, you will at some point add Tegra 3 support I suppose. In > that case much of this code will become common anyway. Should you do > that split (tegra-common, tegra2-common) now or later? I'd like to keep the tegra3 changes separate since that will involve refactoring of much of this code. For the SPL build it's mostly hidden in the config and Makefiles. I thought about doing exactly what you said to reduce the amount of code I needed to pull into the SPL build, but what I found was that because the SPL needs to initialize some regulators, clocks, and pinmux to bring up the A9 and UART it ends up needing much of the tegra2 SoC code, so there's not a lot of opportunity to minize the code set for the SPL build beyond the CONFIG options to disable drivers that area already there. -Allen nvpublic ___ U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot