netdev - Re: [PATCH regression] net: phy: fix initialization (config

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <533D8221.7010103@ahsoftware.de>
Date:	Thu, 03 Apr 2014 17:45:37 +0200
From:	Alexander Holler <holler@...oftware.de>
To:	David Miller <davem@...emloft.net>
CC:	sebastian.hesselbarth@...il.com, f.fainelli@...il.com,
	linux-kernel@...r.kernel.org, netdev@...r.kernel.org,
	michal.simek@...inx.com, stable@...r.kernel.org
Subject: Re: [PATCH regression] net: phy: fix initialization (config_init)
 for Marvel 88E1116R PHYs

Am 03.04.2014 17:14, schrieb David Miller:
> From: Alexander Holler <holler@...oftware.de>
> Date: Thu, 03 Apr 2014 17:06:52 +0200
>
>> But don't suggest me (or insist on) a time consuming
>
> Bisects are not time consuming, and help developers analyze your
> issue tremendously.

Hmm, compiling and booting several dozen kernels isn't time consuming? 
Then I must have done something wrong in the past.

And I wonder why I've writen descriptions at all, if nobody wants to 
understand that I already know the patch which leads to a broken network 
on that system.

If the patch is really the reason for the problem, is something totally 
different, but bisecting won't help here. Besides that I know since 4 
years that netconsole is broken on that box, it just never broke the 
network or anything else. And so I ignored it, foreseeing the necessary 
time to debug and especially describe and discuss the problem.

You see, I'm already feeling like just talking with myself. But just in 
case: After reverting patch 7cd1463, there exist a online-change to turn 
on/off the problem:

------
diff --git a/drivers/net/ethernet/marvell/mv643xx_eth.c 
b/drivers/net/ethernet/marvell/mv643xx_eth.c
index e891b48..246f065 100644
--- a/drivers/net/ethernet/marvell/mv643xx_eth.c
+++ b/drivers/net/ethernet/marvell/mv643xx_eth.c
@@ -2095,7 +2095,8 @@ static void port_start(struct mv643xx_eth_private *mp)
                 struct ethtool_cmd cmd;

                 mv643xx_eth_get_settings(mp->dev, &cmd);
-               phy_reset(mp);
+               //phy_reset(mp);
+               phy_init_hw(mp->phy);
                 mv643xx_eth_set_settings(mp->dev, &cmd);
                 phy_start(mp->phy);
         }
------

Thats basically just what the reverted patch does. Now, why should I 
still bisect?

Of course, my first assumption that the changed reset is the bug is 
wrong, it just makes the problem to appear. But that wrong assumption 
still doesn't make a bisect necessary. And it's discouraging if people 
still insist on that.

I will see if I spend the time to write down a more detailed description 
about what happens here.

Regards,

Alexander Holler
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html