uboot 2024.07 regression

Hi,
while testing clock-patches for mediatek as preparation for OF_upstream i've noticed an issue on mt7623 based board (bananapi r2) with ethernet (no ping possible) in 2024.07-rc1 where 2024.04 is good.
i bisected it to this commit: ff3bcb17f436 2024-02-23 Makefile: pass -undef option to cmd_gen_envp and when reverting this commit in 2024.07 ethernet works
it seems not yet reported or reverted (till 2024.07-rc4).
I'm not sure why this commit breaks network on this board, but revert confirms it...
regards Frank

Hi,
any thoughts about this?
Changes do not look like they would cause the ethernet-breakage, but revert confirms...but i have no idea why this happens (maybe some optimization is triggered).
regards Frank
Gesendet: Samstag, 15. Juni 2024 um 10:35 Uhr Von: "Frank Wunderlich" frank-w@public-files.de An: u-boot@lists.denx.de, "Sébastien Szymanski" sebastien.szymanski@armadeus.com Cc: "Daniel Golle" daniel@makrotopia.org, "John Crispin" john@phrozen.org, "Christian Marangi (Ansuel)" ansuelsmth@gmail.com Betreff: uboot 2024.07 regression
Hi,
while testing clock-patches for mediatek as preparation for OF_upstream i've noticed an issue on mt7623 based board (bananapi r2) with ethernet (no ping possible) in 2024.07-rc1 where 2024.04 is good.
i bisected it to this commit: ff3bcb17f436 2024-02-23 Makefile: pass -undef option to cmd_gen_envp and when reverting this commit in 2024.07 ethernet works
it seems not yet reported or reverted (till 2024.07-rc4).
I'm not sure why this commit breaks network on this board, but revert confirms it...
regards Frank

Hello,
On 6/30/24 16:10, Frank Wunderlich wrote:
Hi,
any thoughts about this?
Do you use an environment file ? If so, can you share it?
Regards,
Changes do not look like they would cause the ethernet-breakage, but revert confirms...but i have no idea why this happens (maybe some optimization is triggered).
regards Frank
Gesendet: Samstag, 15. Juni 2024 um 10:35 Uhr Von: "Frank Wunderlich" frank-w@public-files.de An: u-boot@lists.denx.de, "Sébastien Szymanski" sebastien.szymanski@armadeus.com Cc: "Daniel Golle" daniel@makrotopia.org, "John Crispin" john@phrozen.org, "Christian Marangi (Ansuel)" ansuelsmth@gmail.com Betreff: uboot 2024.07 regression
Hi,
while testing clock-patches for mediatek as preparation for OF_upstream i've noticed an issue on mt7623 based board (bananapi r2) with ethernet (no ping possible) in 2024.07-rc1 where 2024.04 is good.
i bisected it to this commit: ff3bcb17f436 2024-02-23 Makefile: pass -undef option to cmd_gen_envp and when reverting this commit in 2024.07 ethernet works
it seems not yet reported or reverted (till 2024.07-rc4).
I'm not sure why this commit breaks network on this board, but revert confirms it...
regards Frank

Am 1. Juli 2024 10:19:58 MESZ schrieb "Sébastien Szymanski" sebastien.szymanski@armadeus.com:
Hello,
On 6/30/24 16:10, Frank Wunderlich wrote:
Hi,
any thoughts about this?
Do you use an environment file ? If so, can you share it?
Hi,
yes,I use this as builtin env:
https://github.com/frank-w/u-boot/blob/2024-04-bpi/uEnv_r2.txt
But did not notice any scrambling there with printenv. Imho it should only affect serverip and ipaddr
Regards,
Changes do not look like they would cause the ethernet-breakage, but revert confirms...but i have no idea why this happens (maybe some optimization is triggered).
regards Frank
Gesendet: Samstag, 15. Juni 2024 um 10:35 Uhr Von: "Frank Wunderlich" frank-w@public-files.de An: u-boot@lists.denx.de, "Sébastien Szymanski" sebastien.szymanski@armadeus.com Cc: "Daniel Golle" daniel@makrotopia.org, "John Crispin" john@phrozen.org, "Christian Marangi (Ansuel)" ansuelsmth@gmail.com Betreff: uboot 2024.07 regression
Hi,
while testing clock-patches for mediatek as preparation for OF_upstream i've noticed an issue on mt7623 based board (bananapi r2) with ethernet (no ping possible) in 2024.07-rc1 where 2024.04 is good.
i bisected it to this commit: ff3bcb17f436 2024-02-23 Makefile: pass -undef option to cmd_gen_envp and when reverting this commit in 2024.07 ethernet works
it seems not yet reported or reverted (till 2024.07-rc4).
I'm not sure why this commit breaks network on this board, but revert confirms it...
regards Frank
regards Frank

retested with a newer tree (rc4-based) where i have not reverted the commit and there it works again... seems like it as somehow fixed.
sorry for the noise
regards Frank
Gesendet: Montag, 01. Juli 2024 um 10:19 Uhr Von: "Sébastien Szymanski" sebastien.szymanski@armadeus.com An: "Frank Wunderlich" frank-w@public-files.de Cc: u-boot@lists.denx.de, "Daniel Golle" daniel@makrotopia.org, "John Crispin" john@phrozen.org, "Christian Marangi (Ansuel)" ansuelsmth@gmail.com Betreff: Re: Aw: uboot 2024.07 regression
Hello,
On 6/30/24 16:10, Frank Wunderlich wrote:
Hi,
any thoughts about this?
Do you use an environment file ? If so, can you share it?
Regards,
Changes do not look like they would cause the ethernet-breakage, but revert confirms...but i have no idea why this happens (maybe some optimization is triggered).
regards Frank
Gesendet: Samstag, 15. Juni 2024 um 10:35 Uhr Von: "Frank Wunderlich" frank-w@public-files.de An: u-boot@lists.denx.de, "Sébastien Szymanski" sebastien.szymanski@armadeus.com Cc: "Daniel Golle" daniel@makrotopia.org, "John Crispin" john@phrozen.org, "Christian Marangi (Ansuel)" ansuelsmth@gmail.com Betreff: uboot 2024.07 regression
Hi,
while testing clock-patches for mediatek as preparation for OF_upstream i've noticed an issue on mt7623 based board (bananapi r2) with ethernet (no ping possible) in 2024.07-rc1 where 2024.04 is good.
i bisected it to this commit: ff3bcb17f436 2024-02-23 Makefile: pass -undef option to cmd_gen_envp and when reverting this commit in 2024.07 ethernet works
it seems not yet reported or reverted (till 2024.07-rc4).
I'm not sure why this commit breaks network on this board, but revert confirms it...
regards Frank
-- Sébastien Szymanski, Armadeus Systems Software engineer
participants (2)
-
Frank Wunderlich
-
Sébastien Szymanski