[U-Boot] [ARM] Rollback *again* -- 5 patches
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Prafulla (and any other impacted ARM tree custodians), be sure to rebase as necessary.
Stephen, that is an occasion to add your proposal to aad17a2.
Amicalement,
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I am simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I am simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to do it. :)
Amicalement,
Albert,
-----Original Message----- From: Albert ARIBAUD [mailto:albert.u.boot@aribaud.net] Sent: Thursday, April 11, 2013 9:38 AM To: Stephen Warren Cc: U-Boot; Prafulla Wadaskar; Tom Rini; Tom Warren Subject: Re: [ARM] Rollback *again* -- 5 patches
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I am simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to do it. :)
Why should I be fixing something I didn't break, and that worked fine before your change?
As Stephen has pointed out, Tegra uses the whole enchilada (text/rodata/data/bss) for the SPL portion. Any change you add should account for that. Even if we are unique in that use/requirement, it still needs to work correctly before and after your patch.
Tom
Amicalement,
Albert.
----------------------------------------------------------------------------------- This email message is for the sole use of the intended recipient(s) and may contain confidential information. Any unauthorized review, use, disclosure or distribution is prohibited. If you are not the intended recipient, please contact the sender by reply email and destroy all copies of the original message. -----------------------------------------------------------------------------------
Hi Tom,
On Thu, 11 Apr 2013 10:59:29 -0700, Tom Warren TWarren@nvidia.com wrote:
Albert,
-----Original Message----- From: Albert ARIBAUD [mailto:albert.u.boot@aribaud.net] Sent: Thursday, April 11, 2013 9:38 AM To: Stephen Warren Cc: U-Boot; Prafulla Wadaskar; Tom Rini; Tom Warren Subject: Re: [ARM] Rollback *again* -- 5 patches
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I am simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to do it. :)
Why should I be fixing something I didn't break, and that worked fine before your change?
Er...
Was "I did not mean for you to do it" unclear?
Amicalement,
Albert,
On Thu, Apr 11, 2013 at 11:01 AM, Albert ARIBAUD albert.u.boot@aribaud.netwrote:
Hi Tom,
On Thu, 11 Apr 2013 10:59:29 -0700, Tom Warren TWarren@nvidia.com wrote:
Albert,
-----Original Message----- From: Albert ARIBAUD [mailto:albert.u.boot@aribaud.net] Sent: Thursday, April 11, 2013 9:38 AM To: Stephen Warren Cc: U-Boot; Prafulla Wadaskar; Tom Rini; Tom Warren Subject: Re: [ARM] Rollback *again* -- 5 patches
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I
am
simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to
do it. :)
Why should I be fixing something I didn't break, and that worked fine
before your change?
Er...
Was "I did not mean for you to do it" unclear?
Yes, it was unclear.
Tom
Amicalement,
Albert.
Hi Tom,
On Thu, 11 Apr 2013 11:04:44 -0700, Tom Warren twarren.nvidia@gmail.com wrote:
Albert,
On Thu, Apr 11, 2013 at 11:01 AM, Albert ARIBAUD albert.u.boot@aribaud.netwrote:
Hi Tom,
On Thu, 11 Apr 2013 10:59:29 -0700, Tom Warren TWarren@nvidia.com wrote:
Albert,
-----Original Message----- From: Albert ARIBAUD [mailto:albert.u.boot@aribaud.net] Sent: Thursday, April 11, 2013 9:38 AM To: Stephen Warren Cc: U-Boot; Prafulla Wadaskar; Tom Rini; Tom Warren Subject: Re: [ARM] Rollback *again* -- 5 patches
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
...
Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I
am
simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to
do it. :)
Why should I be fixing something I didn't break, and that worked fine
before your change?
Er...
Was "I did not mean for you to do it" unclear?
Yes, it was unclear.
Apologies, then. I only meant to say that we had time to discuss and incorporate your solution in the SPL max size patch series since it had just had to be "unapplied".
Tom
Amicalement,
Hi Albert,
On Thu, 11 Apr 2013 20:13:16 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hi Tom,
On Thu, 11 Apr 2013 11:04:44 -0700, Tom Warren twarren.nvidia@gmail.com wrote:
Albert,
On Thu, Apr 11, 2013 at 11:01 AM, Albert ARIBAUD albert.u.boot@aribaud.netwrote:
Hi Tom,
On Thu, 11 Apr 2013 10:59:29 -0700, Tom Warren TWarren@nvidia.com wrote:
Albert,
-----Original Message----- From: Albert ARIBAUD [mailto:albert.u.boot@aribaud.net] Sent: Thursday, April 11, 2013 9:38 AM To: Stephen Warren Cc: U-Boot; Prafulla Wadaskar; Tom Rini; Tom Warren Subject: Re: [ARM] Rollback *again* -- 5 patches
Hi Stephen,
On Thu, 11 Apr 2013 09:49:45 -0600, Stephen Warren swarren@wwwdotorg.org wrote:
On 04/11/2013 09:25 AM, Albert ARIBAUD wrote: > Hello, > > Older toolchain and PPC incompatibilities appeared in the last two > additions to ToT ARM. To avoid non-bisectability, ToT has been > rolled back from abbecf4c87 to ae74b65987. This removes > > abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts ... > Stephen, that is an occasion to add your proposal to aad17a2.
Tom Warren as the Tegra maintainer should be taking care of this. I
am
simply pointing out that the currently proposed changes to the SPL size checking don't fulfil the requirements for Tegra. I don't expect to have to fix the patches.
I did not mean for you to do it, only that there was an opportunity to
do it. :)
Why should I be fixing something I didn't break, and that worked fine
before your change?
Er...
Was "I did not mean for you to do it" unclear?
Yes, it was unclear.
Apologies, then. I only meant to say that we had time to discuss and incorporate your solution in the SPL max size patch series since it had
*Stephen's* solution.
Sorry, yesterday was kind of hectic.
just had to be "unapplied".
Tom
Amicalement,
Hi Albert,
On Thu, 11 Apr 2013 17:25:34 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Actually, the failure was not due to this series, but to patch 16/30 of Benoît's series. In agreement with Tom, rolling further back, removing the recent TI PR merge and all of Benoît's series.
Benoît, can you fix 16/30 wrt PPC building? If you provide branches, I or Tom can do tests.
Once v12 is ready, I'll apply it back, then the TI PR, then the __bss_end patch, then the SPL max size series, this one list in order to give us time to choose how overall SPL text+data+rodata+BSS size testing should be done.
This still stands:
Prafulla (and any other impacted ARM tree custodians), be sure to rebase as necessary.
Stephen, that is an occasion to add your proposal to aad17a2.
Amicalement,
Amicalement,
Hi Albert,
On Thursday, April 11, 2013 6:30:58 PM, Albert ARIBAUD wrote:
Hi Albert,
On Thu, 11 Apr 2013 17:25:34 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Actually, the failure was not due to this series, but to patch 16/30 of Benoît's series. In agreement with Tom, rolling further back, removing the recent TI PR merge and all of Benoît's series.
Benoît, can you fix 16/30 wrt PPC building? If you provide branches, I or Tom can do tests.
Once v12 is ready, I'll apply it back, then the TI PR, then the __bss_end patch, then the SPL max size series, this one list in order to give us time to choose how overall SPL text+data+rodata+BSS size testing should be done.
OK, I'll do that. I need to know: - which board fails, - where to find the toolchain binaries ready to use for PPC (Ubuntu package?), - if all other arches have been build-tested in order to avoid new issues popping out later.
Best regards, Benoît
Hi Benoît,
On Thu, 11 Apr 2013 18:33:35 +0200 (CEST), Benoît Thébaudeau benoit.thebaudeau@advansee.com wrote:
Hi Albert,
On Thursday, April 11, 2013 6:30:58 PM, Albert ARIBAUD wrote:
Hi Albert,
On Thu, 11 Apr 2013 17:25:34 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Actually, the failure was not due to this series, but to patch 16/30 of Benoît's series. In agreement with Tom, rolling further back, removing the recent TI PR merge and all of Benoît's series.
Benoît, can you fix 16/30 wrt PPC building? If you provide branches, I or Tom can do tests.
Once v12 is ready, I'll apply it back, then the TI PR, then the __bss_end patch, then the SPL max size series, this one list in order to give us time to choose how overall SPL text+data+rodata+BSS size testing should be done.
OK, I'll do that. I need to know:
- which board fails,
P1021RDB-PC_NAND
- where to find the toolchain binaries ready to use for PPC (Ubuntu package?),
Toolchain is ELDK 5.2.1 for powerpc (installs in /opt, need to set PATH=...:/opt/eldk-5.2.1/powerpc/sysroots/i686-eldk-linux/usr/bin/powerpc-linux, ARCH=powerpc, and CROSS_COMPILE=powerpc-linux-
- if all other arches have been build-tested in order to avoid new issues popping out later.
Tom?
Best regards, Benoît
Amicalement,
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 04/11/2013 12:48 PM, Albert ARIBAUD wrote:
Hi Benoît,
On Thu, 11 Apr 2013 18:33:35 +0200 (CEST), Benoît Thébaudeau benoit.thebaudeau@advansee.com wrote:
[snip]
- if all other arches have been build-tested in order to avoid
new issues popping out later.
Tom?
Everything else seemed fine.
- -- Tom
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 04/11/2013 12:33 PM, Benoît Thébaudeau wrote:
Hi Albert,
On Thursday, April 11, 2013 6:30:58 PM, Albert ARIBAUD wrote:
Hi Albert,
On Thu, 11 Apr 2013 17:25:34 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Actually, the failure was not due to this series, but to patch 16/30 of Benoît's series. In agreement with Tom, rolling further back, removing the recent TI PR merge and all of Benoît's series.
Benoît, can you fix 16/30 wrt PPC building? If you provide branches, I or Tom can do tests.
Once v12 is ready, I'll apply it back, then the TI PR, then the __bss_end patch, then the SPL max size series, this one list in order to give us time to choose how overall SPL text+data+rodata+BSS size testing should be done.
OK, I'll do that. I need to know: - which board fails, - where to find the toolchain binaries ready to use for PPC (Ubuntu package?), - if all other arches have been build-tested in order to avoid new issues popping out later.
P1021RDB-PC_NAND fails and I use ELDK 5.2 for a toolchain for PowerPC.
- -- Tom
Hi Albert,
On Thu, 11 Apr 2013 18:30:58 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hi Albert,
On Thu, 11 Apr 2013 17:25:34 +0200, Albert ARIBAUD albert.u.boot@aribaud.net wrote:
Hello,
Older toolchain and PPC incompatibilities appeared in the last two additions to ToT ARM. To avoid non-bisectability, ToT has been rolled back from abbecf4c87 to ae74b65987. This removes
abbecf4 ARM: Fix __bss_start and __bss_end in linker scripts
(HIDDEN not known by most binutils 2.22 or older ones)
aad17a2 ARM: fix CONFIG_SPL_MAX_SIZE semantics 72c1601 smdk5250, snow: fix CONFIG_SPL_MAX_SIZE semantics 55e46fc da850evm, da850_am18xxevm: fix CONFIG_SPL_MAX_SIZE semantics 0502854 cam_enc_4xx: fix CONFIG_SPL_MAX_SIZE semantics
(PPC builds fail because --pad-to receives an expr, not a constant)
Actually, the failure was not due to this series, but to patch 16/30 of Benoît's series. In agreement with Tom, rolling further back, removing the recent TI PR merge and all of Benoît's series.
Status:
- Benoît's series was fixed and build-tested against ARM and PPC.
- Fabio's watchdog removal patch was reapplied.
- Tom's TI PR was re-applied.
ARM Tot (c97b6df1ae92679d67f94c1cfef51323782a506d) can now be used again for basing PRs and patch submissions.
Amicalement,
participants (6)
-
Albert ARIBAUD -
Benoît Thébaudeau -
Stephen Warren -
Tom Rini -
Tom Warren -
Tom Warren