Skip to contentSkip to conversation

net/eth.c: bug-fix

PATCH New Teh Kok How 3 participants 9 messages Feb 12, 2010 – Mar 23, 2010 Networking

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net/eth.c: bug-fix

--- u-boot-2009.11.1.orig/net/eth.c 2010-01-25 16:35:12.000000000 +0800

+++ u-boot-2009.11.1/net/eth.c 2010-02-12 11:18:23.558229895 +0800

@@ -199,7 +199,7 @@

#endif

/* Try board-specific initialization first. If it fails or isn't

* present, try the cpu-specific initialization */

- if (board_eth_init(bis) < 0)

+ if (board_eth_init(bis) <= 0)

                cpu_eth_init(bis);

#if defined(CONFIG_DB64360) || defined(CONFIG_CPCI750)

#1 lore raw

PATCH 1/1 net/eth.c: bug-fix

Quoted 9 lines · Teh Kok How

--- u-boot-2009.11.1.orig/net/eth.c 2010-01-25 16:35:12.000000000 +0800

+++ u-boot-2009.11.1/net/eth.c 2010-02-12 11:18:23.558229895 +0800

(quoted diff) +0 -0
@@ -199,7 +199,7 @@
 
  #endif

Review

Quoted 10 lines
(quoted diff) +1 -1
 
         /* Try board-specific initialization first.  If it fails or isn't
 
          * present, try the cpu-specific initialization */
 
-       if (board_eth_init(bis)<  0)
 
+       if (board_eth_init(bis)<= 0)
 
                 cpu_eth_init(bis);

That's not a bug. The check is meant to indicate whether the weak board_eth_init() function has been overridden, not whether there was an error. I have a clarification patch that hasn't been submitted yet.

regards,
Ben

#2 lore raw

PATCH 1/1 net/eth.c: bug-fix

No, it is a bug. Board_eth_init() calls pci_eth_init() (include/netdev.h) and if ethernet is soc, CONFIG_PCI is not defined, pci_eth_init() returns 0. This patch allows cpu_eth_init() to be called.

-----Original Message-----
From: Ben Warren [mailto:biggerbadderben at gmail.com]
Sent: Friday, February 12, 2010 1:54 PM
To: Teh Kok How
Cc: u-boot at lists.denx.de
Subject: Re: [U-Boot] [PATCH] net/eth.c: bug-fix

Quoted 1 line · Teh Kok How

--- u-boot-2009.11.1.orig/net/eth.c 2010-01-25 16:35:12.000000000

+0800

Quoted 8 lines

+++ u-boot-2009.11.1/net/eth.c 2010-02-12 11:18:23.558229895 +0800

(quoted diff) +0 -0
@@ -199,7 +199,7 @@
 
  #endif

Review

Quoted 10 lines
(quoted diff) +1 -1
 
         /* Try board-specific initialization first.  If it fails or isn't
 
          * present, try the cpu-specific initialization */
 
-       if (board_eth_init(bis)<  0)
 
+       if (board_eth_init(bis)<= 0)
 
                 cpu_eth_init(bis);

That's not a bug. The check is meant to indicate whether the weak board_eth_init() function has been overridden, not whether there was an error. I have a clarification patch that hasn't been submitted yet.

regards,
Ben

#3 lore raw

PATCH 1/1 net/eth.c: bug-fix

Hello Teh Kok How,

Quoted 6 lines · Teh Kok How

No, it is a bug. Board_eth_init() calls pci_eth_init() (include/netdev.h) and if ethernet is soc, CONFIG_PCI is not defined, pci_eth_init() returns 0.
This patch allows cpu_eth_init() to be called.

Sorry, still not a bug. The behavior you've described should be handled

internally to your specific board_eth_init() function. If no PCI devices are found either because none are present or because you didn't specify the CONFIG_PCI option, you can call cpu_eth_init().

regards,
Ben

#4 lore raw

PATCH 1/1 net/eth.c: bug-fix

In reply to Teh Kok How

Fixes: Board_eth_init() calls pci_eth_init() (include/netdev.h) and if ethernet is soc, CONFIG_PCI is not defined, pci_eth_init() returns 0. This patch allows cpu_eth_init() to be called.

net/eth.c 2010-02-12 17:20:32.864273873 +0800 renamed from net/eth.c 2010-01-25 16:35:12.000000000 +0800 +1 -1
@@ -199,7 +199,7 @@
 #endif
        /* Try board-specific initialization first.  If it fails or isn't
         * present, try the cpu-specific initialization */
-       if (board_eth_init(bis) < 0)
+       if (board_eth_init(bis) <= 0)
                cpu_eth_init(bis);
 
 #if defined(CONFIG_DB64360) || defined(CONFIG_CPCI750)
#5 lore raw

PATCH 1/1 net/eth.c: bug-fix

Dear Ben,

Quoted 10 lines · In message <[email protected]> Teh Kok How…

Fixes: Board_eth_init() calls pci_eth_init() (include/netdev.h) and if ethernet is soc, CONFIG_PCI is not defined, pci_eth_init() returns 0. This patch allows cpu_eth_init() to be called.

net/eth.c 2010-02-12 17:20:32.864273873 +0800 line 205

Quoted 8 lines
net/eth.c 2010-02-12 17:20:32.864273873 +0800 renamed from net/eth.c 2010-01-25 16:35:12.000000000 +0800 +1 -1
@@ -199,7 +199,7 @@
 #endif
        /* Try board-specific initialization first.  If it fails or isn't
         * present, try the cpu-specific initialization */
-       if (board_eth_init(bis) < 0)
+       if (board_eth_init(bis) <= 0)
                cpu_eth_init(bis);
 
 #if defined(CONFIG_DB64360) || defined(CONFIG_CPCI750)

Do you have this patch in your queue?

Best regards,

Wolfgang Denk

Signature · 7 lines
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Abstainer: A weak person who yields to the temptation of denying him-
self a pleasure. A total abstainer is one who  abstains  from  every-
thing  but  abstention, and especially from inactivity in the affairs
of others.                                           - Ambrose Bierce
#6 lore raw

PATCH 1/1 net/eth.c: bug-fix

Wolfgang,

Quoted 20 lines · Wolfgang Denk

Dear Ben,

Quoted 16 lines · In message<[email protected]> Teh Kok How…

Fixes: Board_eth_init() calls pci_eth_init() (include/netdev.h) and if ethernet is soc, CONFIG_PCI is not defined, pci_eth_init() returns 0. This patch allows cpu_eth_init() to be called.

net/eth.c 2010-02-12 17:20:32.864273873 +0800 renamed from net/eth.c 2010-01-25 16:35:12.000000000 +0800 +1 -1
@@ -199,7 +199,7 @@
  #endif
         /* Try board-specific initialization first.  If it fails or isn't
          * present, try the cpu-specific initialization */
-       if (board_eth_init(bis)<  0)
+       if (board_eth_init(bis)<= 0)
                 cpu_eth_init(bis);
 
Quoted 6 lines

net/eth.c 2010-02-12 17:20:32.864273873 +0800 line 205

Quoted 3 lines
net/eth.c 2010-02-12 17:20:32.864273873 +0800 renamed from net/eth.c 2010-01-25 16:35:12.000000000 +0800 +0 -0
  #if defined(CONFIG_DB64360) || defined(CONFIG_CPCI750)

Do you have this patch in your queue?

No, I don't. This patch is wrong, and I believe I already responded about it. The code does need a subtle clean-up, but I haven't quite found the time to do it. Hopefully in the next couple of days I'll get to it.

Quoted 4 lines

Best regards,

Wolfgang Denk

regards,
Ben

#7 lore raw

PATCH 1/1 net/eth.c: bug-fix

Dear Ben Warren,

Quoted 7 lines · In message <[email protected]> you wrote:
Quoted 3 lines
Quoted 1 line

Do you have this patch in your queue?

No, I don't. This patch is wrong, and I believe I already responded about it. The code does need a subtle clean-up, but I haven't quite found the time to do it. Hopefully in the next couple of days I'll get to it.

Thanks. I asked because after your reply there was a "But..." from Teh Kok How which AFAICT remained unreplied, so I wa snot sure.

Best regards,

Wolfgang Denk

Signature · 6 lines
DENX Software Engineering GmbH,     MD: Wolfgang Denk & Detlev Zundel
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: (+49)-8142-66989-10 Fax: (+49)-8142-66989-80 Email: wd at denx.de
Punishment becomes ineffective after a certain point. Men become  in-
sensitive.
	-- Eneg, "Patterns of Force", stardate 2534.7
#8 lore raw

PATCH 1/1 net/eth.c: bug-fix

Hi Wolfgang,

Quoted 9 lines · Wolfgang Denk

Dear Ben Warren,

Quoted 5 lines · In message<[email protected]> you wrote:
Quoted 5 lines
Quoted 2 lines

Do you have this patch in your queue?

Quoted 9 lines
Quoted 5 lines

No, I don't. This patch is wrong, and I believe I already responded about it. The code does need a subtle clean-up, but I haven't quite found the time to do it. Hopefully in the next couple of days I'll get to it.

Thanks. I asked because after your reply there was a "But..." from Teh Kok How which AFAICT remained unreplied, so I wa snot sure.

I see that now. Thought I'd responded, but I guess not. Anyway, the right fix for the problem Teh Kok How is (was?) seeing is to have a smarter board_eth_init() function.

Quoted 5 lines

Best regards,

Wolfgang Denk

regards,
Ben