Message ID | 51403C42.6010100@denx.de |
---|---|
State | Rejected |
Delegated to: | Albert ARIBAUD |
Headers | show |
Hi Stefano, On Wed, 13 Mar 2013 09:43:46 +0100, Stefano Babic <sbabic@denx.de> wrote: > Hi Albert, > > please pull from u-boot-imx, thanks. A question about the patch "ARM: > implement some Cortex-A9 errata workarounds" > > http://patchwork.ozlabs.org/patch/223419/ > > I read the general agreement about it, and I have applied 2/3, because > this is i.MX-related. However, the effect is worthless for i.MX6 until > 1/3 will be applied. I read before applying that you will merge it, but > I have not found it in arm-tree. Do you see any problem on it ? If you > do not plan to apply it, it is better I revert 2/3, else the errata will > not be set for i.MX6. Sorry -- I should have taken 1/3 and 2/3 and haven't yet. No worries, though: I'll take them in, then merge your tree; you don't need to revert 2/3, it'll merge properly anyway. > Thanks, > Stefano Amicalement,
On Wed, 13 Mar 2013 13:29:55 +0100, Albert ARIBAUD <albert.u.boot@aribaud.net> wrote: > Hi Stefano, > > On Wed, 13 Mar 2013 09:43:46 +0100, Stefano Babic <sbabic@denx.de> > wrote: > > > Hi Albert, > > > > please pull from u-boot-imx, thanks. A question about the patch "ARM: > > implement some Cortex-A9 errata workarounds" > > > > http://patchwork.ozlabs.org/patch/223419/ > > > > I read the general agreement about it, and I have applied 2/3, because > > this is i.MX-related. However, the effect is worthless for i.MX6 until > > 1/3 will be applied. I read before applying that you will merge it, but > > I have not found it in arm-tree. Do you see any problem on it ? If you > > do not plan to apply it, it is better I revert 2/3, else the errata will > > not be set for i.MX6. > > Sorry -- I should have taken 1/3 and 2/3 and haven't yet. No worries, > though: I'll take them in, then merge your tree; you don't need to > revert 2/3, it'll merge properly anyway. Regardless of applying the two patches above, there is a merge conflict between u-boot-imx/master and u-boot-arm/master between: - commit af73034c6ba131a93ed215098f43595637ef6ac3 ("led: Use STATUS_LED_ON and STATUS_LED_OFF when calling __led_set"), author Otavio Salvador <otavio@ossystems.com.br>, from the IMX tree, and - commit 4f47aceb118b291831a603d7e6ca651d76bf88e0 ("led: The gpio_led.c code expects that LED state is from the enum"), author Holger Hans Peter Freyther <holger@freyther.de>, from the TI tree through the ARM tree. Actually, both commits are functionally exactly the same, Otavio's one being better only in that it causes smaller line widths. I have merged the IMX tree with -Xtheirs so that the conflicts is automatically resolved in favor of Otavio's version. However, MAKEALL -a arm fails on three boards: nitrogen6s1g, with an error, and nitrogen6dl2g and nitrogen6q2g, with a warning. nitrogen6s1g: nitrogen6x.c:89:17: error: 'CONFIG_DDR_MB' undeclared (first use in this function) nitrogen6x.c:89:17: note: each undeclared identifier is reported only once for each function it appears in nitrogen6dl2g and nitrogen6q2g: nitrogen6x.c:89:38: warning: integer overflow in expression [-Woverflow] These failures also happen on IMX ToT before merge. Stefano, can you have a look? I can obviously not locate the actual faulty commit, as support for these boards appears only in the very last commit of the IMX branch. > > Thanks, > > Stefano Amicalement,
Hi Albert, On 03/13/2013 03:26 PM, Albert ARIBAUD wrote: > On Wed, 13 Mar 2013 13:29:55 +0100, Albert ARIBAUD > <albert.u.boot@aribaud.net> wrote: > > <snip> > > However, MAKEALL -a arm fails on three boards: nitrogen6s1g, with an > error, and nitrogen6dl2g and nitrogen6q2g, with a warning. > > nitrogen6s1g: > > nitrogen6x.c:89:17: error: 'CONFIG_DDR_MB' undeclared (first use in > this function) > nitrogen6x.c:89:17: note: each undeclared identifier is reported only > once for each function it appears in > This one's on me. Sorry for not running MAKEALL before submission: The line beginning with nitrogen6s1g should say "DDR_MB=1024", not "SDRAM_MB=1024": http://git.denx.de/u-boot.git/?p=u-boot/u-boot-imx.git;a=blobdiff;f=boards.cfg;h=fb3b197eef192fe5bf5ecaad8bcca5cab3bf1c43;hp=7a0b79dd072f4b8a8c39fe75d638c4d163bdc385;hb=d67b0d97b156f9ec2fc4c838d84b1e510d6e49b4;hpb=7315e3bf2cc95abf9ae53f43ccbd31c6f638aacd > nitrogen6dl2g and nitrogen6q2g: > > nitrogen6x.c:89:38: warning: integer overflow in expression [-Woverflow] > Looking into this now. Regards, Eric
Hi Eric, On Wed, Mar 13, 2013 at 7:36 PM, Eric Nelson <eric.nelson@boundarydevices.com> wrote: > Hi Albert, > > > On 03/13/2013 03:26 PM, Albert ARIBAUD wrote: >> >> On Wed, 13 Mar 2013 13:29:55 +0100, Albert ARIBAUD >> <albert.u.boot@aribaud.net> wrote: >> >> <snip> > >> >> However, MAKEALL -a arm fails on three boards: nitrogen6s1g, with an >> error, and nitrogen6dl2g and nitrogen6q2g, with a warning. >> >> nitrogen6s1g: >> >> nitrogen6x.c:89:17: error: 'CONFIG_DDR_MB' undeclared (first use in >> this function) >> nitrogen6x.c:89:17: note: each undeclared identifier is reported only >> once for each function it appears in >> > > This one's on me. > > Sorry for not running MAKEALL before submission: > > The line beginning with nitrogen6s1g should say "DDR_MB=1024", not > "SDRAM_MB=1024": > > http://git.denx.de/u-boot.git/?p=u-boot/u-boot-imx.git;a=blobdiff;f=boards.cfg;h=fb3b197eef192fe5bf5ecaad8bcca5cab3bf1c43;hp=7a0b79dd072f4b8a8c39fe75d638c4d163bdc385;hb=d67b0d97b156f9ec2fc4c838d84b1e510d6e49b4;hpb=7315e3bf2cc95abf9ae53f43ccbd31c6f638aacd > > >> nitrogen6dl2g and nitrogen6q2g: >> >> nitrogen6x.c:89:38: warning: integer overflow in expression [-Woverflow] >> > > Looking into this now. Just fixed it. Will send the patches now.
On 13/03/2013 23:26, Albert ARIBAUD wrote: > On Wed, 13 Mar 2013 13:29:55 +0100, Albert ARIBAUD > <albert.u.boot@aribaud.net> wrote: > >> Hi Stefano, Hi Albert, > Regardless of applying the two patches above, there is a merge conflict > between u-boot-imx/master and u-boot-arm/master between: > > - commit af73034c6ba131a93ed215098f43595637ef6ac3 ("led: Use > STATUS_LED_ON and STATUS_LED_OFF when calling __led_set"), author > Otavio Salvador <otavio@ossystems.com.br>, from the IMX tree, and > > - commit 4f47aceb118b291831a603d7e6ca651d76bf88e0 ("led: The gpio_led.c > code expects that LED state is from the enum"), author Holger Hans > Peter Freyther <holger@freyther.de>, from the TI tree through the ARM > tree. > > Actually, both commits are functionally exactly the same, Otavio's one > being better only in that it causes smaller line widths. > > I have merged the IMX tree with -Xtheirs so that the conflicts is > automatically resolved in favor of Otavio's version. > Thanks for fixing that. > These failures also happen on IMX ToT before merge. Stefano, can you > have a look? I can obviously not locate the actual faulty commit, as > support for these boards appears only in the very last commit of the IMX > branch. Sorry for that - I have already seen patches from Fabio fixing that. I will test it and then I will send a pull request to fix it. Best regards, Stefano