Stephen,
-----Original Message----- From: Stephen Warren [mailto:swarren@wwwdotorg.org] Sent: Thursday, November 29, 2012 12:10 PM To: Tom Warren Cc: Simon Glass; u-boot@lists.denx.de; Stephen Warren; Marc Dietrich; Thierry Reding Subject: Re: [PATCH 1/3] tegra: only define TEGRA_DEVICE_SETTINGS if not already defined
On 11/29/2012 11:40 AM, Tom Warren wrote:
Stephen,
-----Original Message----- From: Stephen Warren [mailto:swarren@wwwdotorg.org] Sent: Wednesday, November 28, 2012 2:03 PM To: Simon Glass Cc: u-boot@lists.denx.de; Tom Warren; Stephen Warren; Marc Dietrich; Thierry Reding Subject: Re: [PATCH 1/3] tegra: only define TEGRA_DEVICE_SETTINGS if not already defined
On 11/28/2012 02:01 PM, Simon Glass wrote:
Hi Stephen,
On Wed, Nov 28, 2012 at 9:50 AM, Stephen Warren swarren@wwwdotorg.org
wrote:
From: Stephen Warren swarren@nvidia.com
seaboard.h attempts to undefine TEGRA_DEVICE_SETTINGS and provide a custom value. This worked when the "pre" included tegra20-common.h provided the default. However, changes in the main U-Boot repo removed this default from the "pre" included tegra20-common.h to the
"post"
included tegra-common-post.h, which uncondtionally provides the value. This causes the following compile warnings:
In file included from /home/swarren/shared/git_wa/u-
boot/include/configs/seaboard.h:129:0,
from /home/swarren/shared/git_wa/u-boot/include/config.h:10,
from /home/swarren/shared/git_wa/u-boot/include/common.h:37,
from lib/asm-offsets.c:18:/home/swarren/shared/git_wa/u-boot/include/configs/tegra-common-post. h:163:0: warning: "TEGRA_DEVICE_SETTINGS" redefined /home/swarren/shared/git_wa/u-boot/include/configs/seaboard.h:110:0: note: this is the location of the previous definition
Solve this by modifying tegra-common-post.h to only provide a value for TEGRA_DEVICE_SETTINGS if the board-specific header has not already provided a custom value.
Signed-off-by: Stephen Warren swarren@nvidia.com
These series doesn't apply to u-boot-tegra/master or /next for me, and the last one doesn't seem to apply to u-boot/master either. Can you please take a look, may be a timing issue.
Yes, as I mentioned this problem will only exist once u-boot/master and u- boot-arm/master are merged together, so this patch series applies to the result of the merge, which will be (at least part of) the state of u-boot-tegra/* at some unspecified future time:-)
Allen's TEGRA_DEVICE_SETTINGS patch is already in u-boot/master.
Yup.
I applied it to u-boot-tegra/master, then applied your 1/3 & 2/3 patches (3/3 isn't needed if Allen's patch is already in).
Maybe. It depends how Tom Rini does the u-boot-arm/master -> u-boot/master merge; u-boot-arm/master contains a patch that edits TEGRA_DEVICE_SETTINGS in seaboard.h (Simon's to add LCD support), and it's not obvious when merging those two branches that the merge result should be to remove that change from seaboard.h. No actually, that shouldn't be the merge result; if it was, the LCD additions would be removed from TEGRA_DEVICE_SETTINGS without patch 2/3 that I posted. That is, unless patch 2/3 of mine gets into u-boot-arm/master before Albert sends a pull request to Tom Rini...
So, I think that once you rebase onto an upstream branch that itself includes all these patches rather than manually applying them, you will probably find patch 3/3 is required. Admittedly it's not if you just apply Allen's patch to the current u-boot-arm/master.
I'm not anticipating doing a new pull request for u-boot-tegra/master any time real soon, so we can wait for upstream (u-boot-arm & u-boot/master) to settle a bit. But regardless, Albert has said that he's fine with custodians submitting pull requests that contain/depend upon patches that are already in an upstream repo, or with the custodian stating explicitly which upstream patches are needed for a pull request to succeed/build/work. Note that the only patch I'm manually applying per se is Allen's TEGRA_DEVICE_SETTINGS patch that went in to u-boot/master via u-boot-usb via Marek. The rest are true Tegra patches.
Take a look at tegra/master (or /next) and compare to u-boot/master, u-boot-arm, u-boot-usb, etc. if needed and let me know if you still think it's going to run into trouble if/when I do a pull request to arm/master.
Thanks,
Tom
-- nvpublic