Skip to contentSkip to conversation

Ethernet PHY chip discovery not working on 855T with 971/972 chips

DISCUSSION Dean Matsen 10 participants 30 messages Jul 9, 2003 – Jul 15, 2003 1 message not in the archive

lore
Share & tools
lore
#0 lore raw

To whomever maintains the file arch/ppc/8xx_io/fec.c:

I note the following code in mii_discover_phy() :

    fep = dev->priv;

    if ((phytype = (mii_reg & 0xffff)) != 0xffff) {

        /* Got first part of ID, now get remainder.
        */

In our devices, reading the first word with the wrong address yields the value 0, not 0xFFFF. This happens even if we play with it by hand. The search goes through address 0, 1, 2, .., and our chip is hard-coded at address 1. What happens is the kernel panics with something like "Don't know what chip ID 0000FFFF is". Anyway, I changed it to the following, and it finds our PHY correctly now.

    fep = dev->priv;

    phytype = (mii_reg & 0xffff);
    if (phytype != 0xffff && phytype != 0x0000) {

        /* Got first part of ID, now get remainder.
        */

I am guessing that probably most devices have the PHY chip at address zero, so nobody else has run into this problem because it finds their chip in the first iteration.

You may also want to note in the file that it supports 972 chips (which apparently look like 971 chips from the inside, 'cuz that's what the kernel says it is). Perhaps the printk() should be updated to say "971/972" instead of just "971"?

Thanks
Dean Matsen

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#3 lore raw
Quoted 1 line · Eugene Surovegin

It usually means that your hw lacks pull-up on MII MDIO line.

Yep, and even when PHYs indicate they have internal pull-ups, either they don't or they aren't sufficient.

	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#4 lore raw
Quoted 15 lines · Dan Malek
Quoted 1 line · Eugene Surovegin

It usually means that your hw lacks pull-up on MII MDIO line.

Yep, and even when PHYs indicate they have internal pull-ups, either they don't or they aren't sufficient.

-- Dan

Yes, I see we have no pullup on that line. We do dedicated HVAC applications, so we
don't need to do this search. Our dedicated software wouldn't talk to the chip
at the wrong address.

Don't you think the fix from my original message should still be considered for
integration into the kernel? It will improve the odds that the kernel can auto-detect someone's hardware....

Thanks
Dean

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#5 lore raw
Quoted 3 lines · Dean Matsen

Yes, I see we have no pullup on that line. We do dedicated HVAC applications, so we
don't need to do this search.

But.......the hardware is broken. The MDIO is supposed to be an open-drain with a pullup. I guess you may be lucky the ends of the connection are driving the line proplerly, but it could be luck that will run out someday.

	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#6 lore raw
Quoted 8 lines · Dan Malek
Quoted 3 lines · Dean Matsen

Yes, I see we have no pullup on that line. We do dedicated HVAC applications, so we
don't need to do this search.

Quoted 8 lines

But.......the hardware is broken. The MDIO is supposed to be an open-drain with a pullup. I guess you may be lucky the ends of the connection are driving the line proplerly, but it could be luck that will run out someday.

-- Dan

Well, I don't disagree there. We're going to investigate updating the schematics for future products. Thanks for the input.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#7 lore raw
In reply to Dan Malek
Quoted 8 lines · Dan Malek
Quoted 3 lines · Dean Matsen

Yes, I see we have no pullup on that line. We do dedicated HVAC applications, so we
don't need to do this search.

Quoted 7 lines

But.......the hardware is broken. The MDIO is supposed to be an open-drain with a pullup. I guess you may be lucky the ends of the connection are driving the line proplerly, but it could be luck that will run out someday.

    -- Dan

Actually, it turns out that this is not quite true. The MDIO line is classified as an I/O line (on the same page where other lines are classified as open drain). There are tons of application notes from Motorola and Intel and none of them show a pullup. I think this issue has more to do with talking to a device at the wrong address and having the line float only in that case.

The actual problem I have is that the kernel panics before it even gets to address 1 (which is where my PHY chip is in the first place). This happens whenever it reads a value other than FFFF from the first word of the ID. I see now that my first patch adds detection for the value 0000, but other people may have varying results (especially if they follow the app notes and don't have a pullup, like we do).

I think the search should be made smarter (especially, keep going to the next address until it has gone through ALL address w/o finding a valid ID). The kernel could still panic if it can't find any PHY chip, but not the first time it reads the floating line and can't find the answer in its table. The odds of reading a valid PHY ID from the floating line is phenominally low. I'd submit a proposed patch for this, but it doesn't sound like anyone else thinks this is important enough to warrant a change.

Those maintaining the official kernel repositories should consider whether or not it's friendly to require the addition of a pullup resistor just so the automatic PHY search will work. Unless the pullup is required for non-search applications (which I assert it isn't), then I think the code should be fixed, not the hardware.

Regards,

Dean

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#8 lore raw
Quoted 9 lines · In message <[email protected]> you wrote:

I think the search should be made smarter (especially, keep going to the next address until it has gone through ALL address w/o finding a valid ID). The kernel could still panic if it can't find any PHY chip, but not the first time it reads the floating line and can't find the answer in its table. The odds of reading a valid PHY ID from the floating line is phenominally low. I'd submit a proposed patch for this, but it doesn't sound like anyone else thinks this is important enough to warrant a change.

I agree that it makes little sense for th kernel to panic if it finds an unsupported PHY id.

Quoted 3 lines

the automatic PHY search will work. Unless the pullup is required for non-search applications (which I assert it isn't), then I think the code should be fixed, not the hardware.

We changed this in our kernel tree some time ago. If no valid PHY is found we simply do not enable the interface, but we do not panic any more.

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected]
"A witty saying proves nothing." - Voltaire

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

A message that is not in the archive (sent off-list or not yet received): [email protected]

#10 lore raw

Dear Tom,

Quoted 8 lines · in message <[email protected]> you wrote:
Quoted 4 lines

We changed this in our kernel tree some time ago. If no valid PHY is found we simply do not enable the interface, but we do not panic any more.

Which raises two fun questions:
1) Do you plan on ever trying to re-sync your 2.4 tree with the community tree (which should become kernel.org before too long)?

We do this from time to time. The last (visible) re-sync was with 2.4.20 (ChangeSet 1.1174 = 20021129181904|49831), which is half a year ago. On the other hand, we don't have the resources to follow each and every change immediately.

Quoted 2 lines

2) Do you plan on trying to keep in sync with 2.6 now that it's coming up on us?

Of course we will follow 2.6 as well.

Why are you asking? Did I give any reason to think we might go out of business?

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] "There are three principal ways to lose money: wine, women, and en- gineers. While the first two are more pleasant, the third is by far
the more certain." -- Baron Rothschild, ca. 1800

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#11 lore raw
Quoted 17 lines · Wolfgang Denk

Dear Tom,

Quoted 8 lines · in message <[email protected]> you wrote:
Quoted 4 lines

We changed this in our kernel tree some time ago. If no valid PHY is found we simply do not enable the interface, but we do not panic any more.

Which raises two fun questions:
1) Do you plan on ever trying to re-sync your 2.4 tree with the community tree (which should become kernel.org before too long)?

We do this from time to time. The last (visible) re-sync was with 2.4.20 (ChangeSet 1.1174 = 20021129181904|49831), which is half a year ago. On the other hand, we don't have the resources to follow each and every change immediately.

Quoted 7 lines
Quoted 2 lines

2) Do you plan on trying to keep in sync with 2.6 now that it's coming up on us?

Of course we will follow 2.6 as well.

Why are you asking? Did I give any reason to think we might go out of business?

No, but I do wonder if you ever plan on sending anything out again. It does seem to be a once a year event, which is just about due (2002-06-18 is the last one I see).

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#12 lore raw

Dear Tom,

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

No, but I do wonder if you ever plan on sending anything out again. It

Has anything significantly changed? Patches/extensions are still accepted for 2.5 only, right? Does this tree work for 8xx systems?

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times. Or see thread "set_rtc_time() cleanup / normalization" in May this year.

If you suggest there is a way to get patches really included into the "official" trees, please let me know how to do this.

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] You see things; and you say ``Why?'' But I dream things that never were; and I say ``Why not?''
- George Bernard Shaw _Back to Methuselah_ (1921) pt. 1, act 1

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#13 lore raw
Quoted 9 lines · Wolfgang Denk

Dear Tom,

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

No, but I do wonder if you ever plan on sending anything out again. It

Has anything significantly changed? Patches/extensions are still accepted for 2.5 only, right? Does this tree work for 8xx systems?

2.5 untested, so that when 2.6 rolls around there is less pain for all, and 2.4 tested. This hasn't changed, and is becoming a bit more of a trend even outside of the PPC world. It's not even that hard to test now, doing 'make ARCH=ppc CROSS_COMPILE=foo arch/ppc/foo/bar.o' will work and is a feature of the new build system.

Quoted 4 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

Quoted 2 lines

Or see thread "set_rtc_time() cleanup / normalization" in May this year.

I just re-read most of it, and Gabriel Paubert, who wrote most / all of that code, and is the 'maintainer' of sorts objected, a few solutions to make everyone happy were proposed (ppc_md hook or similar, I believe) and no further patch was submitted.

Quoted 2 lines

If you suggest there is a way to get patches really included into the "official" trees, please let me know how to do this.

As always, send things in functional hunks. If for example, the 8xx / 8260 enet and serial drivers (Of note: the 4xx enet driver now falls into this catagory) in 2.6 get moved into drivers/serial (And rewritten) and drivers/net (cleaned up?), rmk (serial maintainer) and Jeff Garzik (net driver maintainer) aren't going to take a large patch without at least as much griping as I've made in the past, if not more. And when someone comments on the reasoning, logic, or suggests an alternate way of doing things for people which need the current behavior / whatever, re-spin the patch. All of the patches you've submitted to me like this, AFAIK, have either been accepted, or I've told you I cannot take that patch, it needs to be sent elsewhere.

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#14 lore raw
Quoted 6 lines · In message <[email protected]> you wrote:

2.5 untested, so that when 2.6 rolls around there is less pain for all, and 2.4 tested. This hasn't changed, and is becoming a bit more of a trend even outside of the PPC world. It's not even that hard to test now, doing 'make ARCH=ppc CROSS_COMPILE=foo arch/ppc/foo/bar.o' will work and is a feature of the new build system.

Does not work for me; even simple core files fail to compile:

-> make ARCH=ppc CROSS_COMPILE=ppc_8xx- arch/ppc/syslib/ppc8xx_pic.o
CC arch/ppc/syslib/ppc8xx_pic.o
arch/ppc/syslib/ppc8xx_pic.c:172: parse error before irqreturn_t' arch/ppc/syslib/ppc8xx_pic.c:172: request_irq' declared as function returning a function arch/ppc/syslib/ppc8xx_pic.c:172: warning: function declaration isn't a prototype arch/ppc/syslib/ppc8xx_pic.c:173: parse error before `unsigned'
make[1]: *** [arch/ppc/syslib/ppc8xx_pic.o] Error 1
make: *** [arch/ppc/syslib/ppc8xx_pic.o] Error 2

Quoted 6 lines
Quoted 3 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

Sorry, I don't have that much time to go thtough this again. AFAIR

Quoted 7 lines
Quoted 2 lines

Or see thread "set_rtc_time() cleanup / normalization" in May this year.

I just re-read most of it, and Gabriel Paubert, who wrote most / all of that code, and is the 'maintainer' of sorts objected, a few solutions to make everyone happy were proposed (ppc_md hook or similar, I believe) and no further patch was submitted.

IIRC there was a general disagreement - what patch do you expect then?

Quoted 3 lines

As always, send things in functional hunks. If for example, the 8xx / 8260 enet and serial drivers (Of note: the 4xx enet driver now falls into this catagory) in 2.6 get moved into drivers/serial (And rewritten)

Please explain what "get moved" means. Is this planned but not started yet? work in progress? who is working on it? who is discussing these things? where does such discussion take place? I've seen nothing like this on lkml? what must one do to participate in such discussions?

Quoted 4 lines

least as much griping as I've made in the past, if not more. And when someone comments on the reasoning, logic, or suggests an alternate way of doing things for people which need the current behavior / whatever, re-spin the patch. All of the patches you've submitted to me like this,

Sorry, but how should anybody comment before being faced with accomplished facts?

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] Hegel was right when he said that we learn from history that man can
never learn anything from history. - George Bernard Shaw

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#15 lore raw
Quoted 10 lines · Wolfgang Denk
Quoted 6 lines · In message <[email protected]> you wrote:

2.5 untested, so that when 2.6 rolls around there is less pain for all, and 2.4 tested. This hasn't changed, and is becoming a bit more of a trend even outside of the PPC world. It's not even that hard to test now, doing 'make ARCH=ppc CROSS_COMPILE=foo arch/ppc/foo/bar.o' will work and is a feature of the new build system.

Does not work for me; even simple core files fail to compile:

Quoted 8 lines

-> make ARCH=ppc CROSS_COMPILE=ppc_8xx- arch/ppc/syslib/ppc8xx_pic.o
CC arch/ppc/syslib/ppc8xx_pic.o
arch/ppc/syslib/ppc8xx_pic.c:172: parse error before irqreturn_t' arch/ppc/syslib/ppc8xx_pic.c:172: request_irq' declared as function returning a function arch/ppc/syslib/ppc8xx_pic.c:172: warning: function declaration isn't a prototype arch/ppc/syslib/ppc8xx_pic.c:173: parse error before `unsigned'
make[1]: *** [arch/ppc/syslib/ppc8xx_pic.o] Error 1
make: *** [arch/ppc/syslib/ppc8xx_pic.o] Error 2

Are you modifying ppc8xx_pic.c ? And yes, it does need generic updates.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

Sorry, I don't have that much time to go thtough this again. AFAIR

You don't have time to find a pointer to a discussion? Then I can't say why nothing happened, but I'll wager a guess that Dan Malek didn't like it for some reason.

Quoted 9 lines
Quoted 7 lines
Quoted 2 lines

Or see thread "set_rtc_time() cleanup / normalization" in May this year.

I just re-read most of it, and Gabriel Paubert, who wrote most / all of that code, and is the 'maintainer' of sorts objected, a few solutions to make everyone happy were proposed (ppc_md hook or similar, I believe) and no further patch was submitted.

IIRC there was a general disagreement - what patch do you expect then?

If there was a general disagreement (I thought the add a do or don't do this in an interrupt flag was accepted) I can't go and apply the patch anyhow. That's also not something I can help you with[1].

Quoted 9 lines
Quoted 3 lines

As always, send things in functional hunks. If for example, the 8xx / 8260 enet and serial drivers (Of note: the 4xx enet driver now falls into this catagory) in 2.6 get moved into drivers/serial (And rewritten)

Please explain what "get moved" means. Is this planned but not started yet? work in progress? who is working on it? who is discussing these things? where does such discussion take place? I've seen nothing like this on lkml? what must one do to participate in such discussions?

A hypotheical. Not that the serial drivers shouldn't be moved to the new serial infrastructure that'll c&p bugs...

Quoted 7 lines
Quoted 4 lines

least as much griping as I've made in the past, if not more. And when someone comments on the reasoning, logic, or suggests an alternate way of doing things for people which need the current behavior / whatever, re-spin the patch. All of the patches you've submitted to me like this,

Sorry, but how should anybody comment before being faced with accomplished facts?

Can you rephrase that? What I was saying is that frequently Paul has pointed to something done outside of the 8xx / 8260 world and asked if we can try doing it in a different way instead, which he believes is cleaner / better / whatever. And since Paul is the PPC maintainer, I (or whomever has been asked) will give it a shot, or explain why that won't work.

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#16 lore raw
Quoted 2 lines · In message <[email protected]> you wrote:

Are you modifying ppc8xx_pic.c ? And yes, it does need generic updates.

No, not at the moment. But how can I understand the stuff in 2.5 if I cannot even compile the code for any standard 8xx system? Any effort to add more code of similar quality is a waste of time.

Quoted 3 lines

You don't have time to find a pointer to a discussion? Then I can't say why nothing happened, but I'll wager a guess that Dan Malek didn't like it for some reason.

I can dig out the original messages when I sent you the patches, I can try to find traces in the archives. It takes time, but what will be the result? If nothing happened then, I don't expect anything to happen now.

You are right, patches get rejected because somebody didn't like it for some reason. But if these patches are addressing reoccuring problems, and no other (better) solution gets implemented, which sense does it make to reject the stuff?

I'm sorry, but both you and me remember that we had discussions like this before, and obviously nothing has changed since. So the result will be the same. We can stop that here.

Quoted 8 lines
Quoted 5 lines

Please explain what "get moved" means. Is this planned but not started yet? work in progress? who is working on it? who is discussing these things? where does such discussion take place? I've seen nothing like this on lkml? what must one do to participate in such discussions?

A hypotheical. Not that the serial drivers shouldn't be moved to the new serial infrastructure that'll c&p bugs...

Let me repeat: who is discussing these things? where does such discussion take place? I've seen nothing like this on lkml? what must one do to participate in such discussions?

Quoted 2 lines

Can you rephrase that? What I was saying is that frequently Paul has pointed to something done outside of the 8xx / 8260 world and asked if

Where has this frequent discussion taken place? I don't remember I've seen it here on the list, nor on lkml (well, I might have missed it on lkml).

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] In the beginning there was nothing.
And the Lord said "Let There Be Light!"
And still there was nothing, but at least now you could see it.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#17 lore raw
In reply to Tom Rini
Quoted 2 lines · Tom Rini

On Fri, Jul 11, 2003 at 08:29:16PM +0200, Wolfgang Denk wrote:

[snip]

Quoted 10 lines
Quoted 5 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

I was the one that sent the patches (search for dpalloc).

A link to the thread in the archive is here.

http://lists.linuxppc.org/linuxppc-embedded/200212/msg00302.html

As it stands I maintain them for the linuxppc_2_4_devel tree and I'm waiting for the 2.5 to build in order to port them there and resubmit.

Quoted 9 lines

--
Tom Rini
http://gate.crashing.org/~trini/

Regards

Pantelis

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#18 lore raw

Dear Pantelis,

Quoted 10 lines · in message <[email protected]> you wrote:
Quoted 9 lines
Quoted 4 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

I was the one that sent the patches (search for dpalloc).

I was referring to a much older implementation of this feature. I sent it to Tom nearly two years ago:

Subject: Re: Modifications to linuxppc_2_4_devel
From: Wolfgang Denk <[email protected]>
Date: Wed, 29 Aug 2001 23:57:30 +0200
To: Tom Rini <[email protected]>

The patch in question was named "PATCH-006-dpfree+uart".

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] Ever try. Ever fail. No matter. Try again. Fail again. Fail better.
-- S. Beckett

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#19 lore raw
Quoted 17 lines · Wolfgang Denk

Dear Pantelis,

Quoted 10 lines · in message <[email protected]> you wrote:
Quoted 9 lines
Quoted 4 lines

IIRC the last time we such discussions simply nothing happened: see for example to code for dp_alloc / dp_free that was submitted several times.

google knows nothing of this, can you give a pointer into the thread please? Thanks.

I was the one that sent the patches (search for dpalloc).

I was referring to a much older implementation of this feature. I sent it to Tom nearly two years ago:

Quoted 6 lines

Subject: Re: Modifications to linuxppc_2_4_devel
From: Wolfgang Denk <[email protected]>
Date: Wed, 29 Aug 2001 23:57:30 +0200
To: Tom Rini <[email protected]>

The patch in question was named "PATCH-006-dpfree+uart".

I really wish I had looked into making mutt keep a copy of sent mail a lot sooner. I don't suppose you have a copy of my reply to that as well? I suspect it was either "Can you break the uart stuff out from the dpfree stuff" and you didn't, or "This is large and intrusive and I'd feel better if I could make Dan Malek look at it".

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#20 lore raw

Dear Tom,

Quoted 3 lines · in message <[email protected]> you wrote:
Quoted 3 lines

I was referring to a much older implementation of this feature. I sent it to Tom nearly two years ago:

...

Quoted 5 lines

I really wish I had looked into making mutt keep a copy of sent mail a lot sooner. I don't suppose you have a copy of my reply to that as well? I suspect it was either "Can you break the uart stuff out from the dpfree stuff" and you didn't, or "This is large and intrusive and I'd feel better if I could make Dan Malek look at it".

I found two related replies in my archive:

28 Aug 2001:
...
PATCH-006 - Looks good. I'll play w/ it a bit more tomorrow and let you
know.
...

15 Oct 2001:
...
PATCH-006 Paul Mundt has been a bit busy of late (and had his disks die a week or so ago), but he should have something soon I hope. Sorry.
...

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected] Der Irrtum wiederholt sich immerfort in der Tat. Deshalb muß man das
Wahre unermüdlich in Worten wiederholen. - Goethe

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#21 lore raw
Quoted 21 lines · Wolfgang Denk

Dear Tom,

Quoted 3 lines · in message <[email protected]> you wrote:
Quoted 3 lines

I was referring to a much older implementation of this feature. I sent it to Tom nearly two years ago:

...

Quoted 5 lines

I really wish I had looked into making mutt keep a copy of sent mail a lot sooner. I don't suppose you have a copy of my reply to that as well? I suspect it was either "Can you break the uart stuff out from the dpfree stuff" and you didn't, or "This is large and intrusive and I'd feel better if I could make Dan Malek look at it".

I found two related replies in my archive:

28 Aug 2001:
...
PATCH-006 - Looks good. I'll play w/ it a bit more tomorrow and let you
know.
...

Quoted 5 lines

15 Oct 2001:
...
PATCH-006 Paul Mundt has been a bit busy of late (and had his disks die a week or so ago), but he should have something soon I hope. Sorry.
...

I found that one (I do keep replies) and that was for an rpxfb patch.

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#22 lore raw
In reply to a message not in the archive
Quoted 39 lines · Tom Rini
Quoted 18 lines · Wolfgang Denk
Quoted 10 lines · In message <[email protected]> you wrote:

I think the search should be made smarter (especially, keep going to the next address until it has gone through ALL address w/o finding a valid ID). The kernel could still panic if it can't find any PHY chip, but not the first time it reads the floating line and can't find the answer in its table. The odds of reading a valid PHY ID from the floating line is phenominally low. I'd submit a proposed patch for this, but it doesn't sound like anyone else thinks this is important enough to warrant a change.

I agree that it makes little sense for th kernel to panic if it finds an unsupported PHY id.

Quoted 10 lines
Quoted 5 lines

the automatic PHY search will work. Unless the pullup is required for non-search applications (which I assert it isn't), then I think the code should be fixed, not the hardware.

We changed this in our kernel tree some time ago. If no valid PHY is found we simply do not enable the interface, but we do not panic any more.

Which raises two fun questions:
1) Do you plan on ever trying to re-sync your 2.4 tree with the community tree (which should become kernel.org before too long)?
2) Do you plan on trying to keep in sync with 2.6 now that it's coming up on us?

Quoted 8 lines

--
Tom Rini
http://gate.crashing.org/~trini/

Somewhere in this discussion thread I saw that the 2.4 kernel is closed for patches (I just joined this list, so I didn't know that).

Would it help if I came up with a 2.5 patch for fec.c? I made a few minor changes that will make it behave friendlier for all [to whom it is not already being totally friendly], and I see the 2.5 kernel could use some of the same changes.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#23 lore raw
Quoted 2 lines · Dean Matsen

Somewhere in this discussion thread I saw that the 2.4 kernel is closed for patches (I just joined this list, so I didn't know that).

Well, that's not strictly true, we will continue to fix things that are broken, but I would prefer than any new development (e.g. support for new boards or devices) happened in 2.5. It's a bit disappointing that 8xx still doesn't compile in 2.5. I don't have any 8xx hardware to test on so I have been relying on others to update it.

Quoted 4 lines

Would it help if I came up with a 2.5 patch for fec.c? I made a few minor changes that will make it behave friendlier for all [to whom it is not already being totally friendly], and I see the 2.5 kernel could use some of the same changes.

Excellent idea. :)

Paul.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#24 lore raw
Quoted 9 lines · Paul Mackerras
Quoted 4 lines · Dean Matsen

Somewhere in this discussion thread I saw that the 2.4 kernel is closed for patches (I just joined this list, so I didn't know that).

Quoted 7 lines

Well, that's not strictly true, we will continue to fix things that are broken, but I would prefer than any new development (e.g. support for new boards or devices) happened in 2.5. It's a bit disappointing that 8xx still doesn't compile in 2.5. I don't have any 8xx hardware to test on so I have been relying on others to update it.

Ok, in that case I do have ONE thing that I know should be fixed, but I haven't seen any response on it, so I am wondering if anyone noticed it at all. See

http://lists.linuxppc.org/linuxppc-dev/200307/msg00085.html

This affects existing support. It's so simple that:
a) It's kind of annoying to 8xx people
b) probably everyone will figure it out
c) there is really no risk in fixing it.

Quoted 9 lines
Quoted 8 lines

Would it help if I came up with a 2.5 patch for fec.c? I made a few minor changes that will make it behave friendlier for all [to whom it is not already being totally friendly], and I see the 2.5 kernel could use some of the same changes.

Quoted 8 lines

Excellent idea. :)

Paul.

I would love to see the kernel perform well for mpc8xx users right out of the
box, so I am highly motivated to contibute. I'll focus on the 2.5 kernel (I'm surprised to hear it doesn't compile).

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#25 lore raw
Quoted 3 lines · In message <[email protected]> you wrote:

box, so I am highly motivated to contibute. I'll focus on the 2.5 kernel (I'm surprised to hear it doesn't compile).

I think it never did. At least never all the times I tried.

Best regards,

Wolfgang Denk

--
Software Engineering: Embedded and Realtime Systems, Embedded Linux
Phone: (+49)-8142-4596-87 Fax: (+49)-8142-4596-88 Email: [email protected]
Administration: An ingenious abstraction in politics, designed to receive the kicks and cuffs due to the premier or president.
- Ambrose Bierce

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#26 lore raw
In reply to Dean Matsen
Quoted 19 lines · Dean Matsen
Quoted 10 lines · Paul Mackerras
Quoted 5 lines · Dean Matsen

Somewhere in this discussion thread I saw that the 2.4 kernel is closed for
patches (I just joined this list, so I didn't know that).

Quoted 7 lines

Well, that's not strictly true, we will continue to fix things that are broken, but I would prefer than any new development (e.g. support for new boards or devices) happened in 2.5. It's a bit disappointing that 8xx still doesn't compile in 2.5. I don't have any 8xx hardware to test on so I have been relying on others to update it.

Quoted 5 lines

Ok, in that case I do have ONE thing that I know should be fixed, but I haven't seen any response on it, so I am wondering if anyone noticed it at all. See

http://lists.linuxppc.org/linuxppc-dev/200307/msg00085.html

This has been fixed in the 2_4_devel tree now. This is not an issue in 2.5 / 2.6.

--
Tom Rini
http://gate.crashing.org/~trini/

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#27 lore raw
In reply to Paul Mackerras
Quoted 3 lines · Paul Mackerras

.... It's a bit disappointing
that 8xx still doesn't compile in 2.5. I don't have any 8xx hardware to test on so I have been relying on others to update it.

Ok, OK....I'll work on it some more. I'm travelling for the next week or so, but when I get back I'll push some updates and get it going.

	-- Dan

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#28 lore raw

Paul> .... It's a bit disappointing that 8xx still doesn't compile Paul> in 2.5. I don't have any 8xx hardware to test on so I have Paul> been relying on others to update it.

Speaking of 2.5, since our custom boards are in production now, I have an Ebony (440GP eval board) that is completely idle, and some free hacking time. Is there any 4xx work that I could help with? Is linuxppc-2.5 the right tree to use?

Best,
Roland

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#29 lore raw
Quoted 8 lines · Roland Dreier

Paul> .... It's a bit disappointing that 8xx still doesn't compile Paul> in 2.5. I don't have any 8xx hardware to test on so I have Paul> been relying on others to update it.

Speaking of 2.5, since our custom boards are in production now, I have an Ebony (440GP eval board) that is completely idle, and some free hacking time. Is there any 4xx work that I could help with? Is

A 44x thing that would be nice is to abstract the PCI-X bridge support into a common call (or calls) so it can be configured into a variety of windowing modes...like alternative monarch memory maps or any non-monarch configurations. Part of this would work would require that fixup_bigphys_addr() be made aware of these window changes. In addition it would be nice to have a helper function that sets the same values as the current hardcoded monarch configuration via the new abstracted interface since that is a common configuration.

Sorry, but you asked. :)

Quoted 1 line

linuxppc-2.5 the right tree to use?

Yes, nothing 44x-related should be going in without a 2.5 patch. I've invested considerable time making sure that it was usable by other developers on 44x...hopefully that still the case after two weeks of vacation.

Regards,
--
Matt Porter
[email protected]

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/

#30 lore raw

Matt> A 44x thing that would be nice is to abstract the PCI-X Matt> bridge support into a common call (or calls) so it can be Matt> configured into a variety of windowing modes...like Matt> alternative monarch memory maps or any non-monarch Matt> configurations. Part of this would work would require that Matt> fixup_bigphys_addr() be made aware of these window changes. Matt> In addition it would be nice to have a helper function that Matt> sets the same values as the current hardcoded monarch Matt> configuration via the new abstracted interface since that is Matt> a common configuration.

That sounds doable, although I'm not sure how to test any non-monarch support given the systems I have available. In any case I'll start taking a look at the current 2.5 sources and try to come up with a proposal for the right PCI-X abstraction.

- R.

** Sent via the linuxppc-embedded mail list. See http://lists.linuxppc.org/