Any value in having a netdev FAQ?
Subsystems (2)
Hi Dave,
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Thanks,
Paul.
--
Information you need to know about netdev
-----------------------------------------
Q: What is netdev?
A: It is a mailing list for all network related linux stuff. This includes
anything found under net/ (i.e. core code like IPv6) and drivers/net (i.e. hardware specific drivers) in the linux source tree.
Note that some subsystems (e.g. wireless drivers) which have a high volume of traffic have their own specific mailing lists.
The netdev list is managed (like many other linux mailing lists) through VGER ( http://vger.kernel.org/ ) and archives can be found below:
http://marc.info/?l=linux-netdev
http://www.spinics.net/lists/netdev/
Aside from subsystems like that mentioned above, all network related linux development (i.e. RFC, review, comments, etc) takes place on netdev.
Q: How do the changes posted to netdev make their way into linux?
A: There are always two trees (git repositories) in play. Both are driven
by David Miller, the main network maintainer. There is the "net" tree, and the "net-next" tree. As you can probably guess from the names, the net tree is for fixes to existing code already in the mainline tree from Linus, and net-next is where the new code goes for the future release. You can find the trees here:
http://git.kernel.org/?p=linux/kernel/git/davem/net.git
http://git.kernel.org/?p=linux/kernel/git/davem/net-next.git
Q: How often do changes from these trees make it to the mainline Linus tree?
A: To understand this, you need to know a bit of background information
on the cadence of linux development. Each new release starts off with a two week "merge window" where the main maintainers feed their new stuff to Linus for merging into the mainline tree. After the two weeks, the merge window is closed, and it is called/tagged "-rc1". No new features get mainlined after this -- only fixes to the rc1 content are expected. After roughly a week of collecting fixes to the rc1 content, rc2 is released. This repeats on a roughly weekly basis until rc7 (typically; sometimes rc6 if things are quiet, or rc8 if things are in a state of churn), and a week after the last vX.Y-rcN was done, the official "vX.Y" is released.
Relating that to netdev: At the beginning of the 2wk merge window, the net-next tree will be closed - no new changes/features. The accumulated new content of the past ~10 weeks will be passed onto mainline/Linus via a pull request for vX.Y -- at the same time, the "net" tree will start accumulating fixes for this pulled content relating to vX.Y
An announcement indicating when net-next has been closed is usually sent to netdev, but knowing the above, you can predict that in advance.
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
The "net" tree continues to collect fixes for the vX.Y content, and is fed back to Linus at regular (~weekly) intervals. Meaning that the focus for "net" is on stablilization and bugfixes.
Finally, the vX.Y gets released, and the whole cycle starts over.
Q: So where are we now in this cycle?
A: Load the mainline (Linus) page here:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git
and note the top of the "tags" section. If it is rc1, it is early in the dev cycle. If it was tagged rc7 a week ago, then a release is probably imminent.
Q: How do I indicate which tree (net vs. net-next) my patch should be in?
A: Firstly, think whether you have a bug fix or new "next-like" content.
Then once decided, assuming that you use git, use the prefix flag, i.e.
git format-patch --subject-prefix='PATCH net-next' start..finish
Use "net" instead of "net-next" in the above for bug-fix net content. If you don't use git, then note the only magic in the above is just the subject text of the outgoing e-mail, and you can manually change it yourself with whatever MUA you are comfortable with.
Q: I sent a patch and I'm wondering what happened to it. How can I tell
whether it got merged?
A: Start by looking at the main patchworks queue for netdev:
http://patchwork.ozlabs.org/project/netdev/list/
The "State" field will tell you exactly where things are at with your patch.
Q: The above only says "Under Review". How can I find out more?
A: Generally speaking, the patches get triaged quickly (in less than 48h).
So be patient. Asking the maintainer for status updates on your patch is a good way to ensure your patch is ignored or pushed to the bottom of the priority list.
Q: How can I tell what patches are queued up for backporting to the
various stable releases?
A: Normally Greg Kroah-Hartman collects stable commits himself, but
for networking, Dave collects up patches he deems critical for the networking subsystem, and then hands them off to Greg.
There is a patchworks queue that you can see here:
http://patchwork.ozlabs.org/bundle/davem/stable/?state=*
It contains the patches which Dave has selected, but not yet handed off to Greg. If Greg already has the patch, then it will be here:
http://git.kernel.org/cgit/linux/kernel/git/stable/stable-queue.git
A quick way to find whether the patch is in this stable-queue is to simply clone the repo, and then git grep the mainline commit ID, e.g.
stable-queue$ git grep -l 284041ef21fdf2e releases/3.0.84/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.4.51/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.9.8/ipv6-fix-possible-crashes-in-ip6_cork_release.patch stable/stable-queue$
Q: I see a network patch and I think it should be backported to stable.
Should I request it via "[email protected]" like the references in the kernel's Documentation/ directory say?
A: No, not for networking. Check the stable queues as per above 1st to see
if it is already queued. If not, then send a mail to netdev, listing the upstream commit ID and why you think it should be a stable candidate.
Before you jump to go do the above, do note that the normal stable rules in Documentation/stable_kernel_rules.txt still apply. So you need to explicitly indicate why it is a critical fix and exactly what users are impacted. In addition, you need to convince yourself that you _really_ think it has been overlooked, vs. having been considered and rejected.
Generally speaking, the longer it has had a chance to "soak" in mainline, the better the odds that it is an OK candidate for stable. So scrambling to request a commit be added the day after it appears should be avoided.
Q: I have created a network patch and I think it should be backported to
stable. Should I add a "Cc: [email protected]" like the references in the kernel's Documentation/ directory say?
A: No. See above answer. In short, if you think it really belongs in
stable, then ensure you write a decent commit log that describes who gets impacted by the bugfix and how it manifests itself, and when the bug was introduced. If you do that properly, then the commit will get handled appropriately and most likely get put in the patchworks stable queue if it really warrants it.
If you think there is some valid information relating to it being in stable that does _not_ belong in the commit log, then use the three dash marker line as described in Documentation/SubmittingPatches to temporarily embed that information into the patch that you send.
Q: Someone said that the comment style and coding convention is different
for the networing content. Is this true?
A: Yes, in a largely trivial way. Instead of this:
/* * foobar blah blah blah * another line of text */
it is requested that you make it look like this:
/* foobar blah blah blah * another line of text */
Q: I am working in existing code that has the former comment style and not the
latter. Should I submit new code in the former style or the latter?
A: Make it the latter style, so that eventually all code in the domain of
netdev is of this format.
Q: I found a bug that might have possible security implications or similar.
Should I mail the main netdev maintainer off-list?
A: No. The current netdev maintainer has consistently requested that people
use the mailing lists and not reach out directly. If you aren't OK with that, then perhaps consider using "[email protected]" instead.
Q: What level of testing is expected before I submit my change?
A: If your changes are against net-next, then the expectation is that
you have tested by layering your changes on top of net-next. Ideally you will have done run-time testing specific to your change, but at a minimum, your changes should survive an "allyesconfig" and an "allmodconfig" build without new warnings or failures.
Q: Any other tips to help ensure my net/net-next patch gets OK'd?
A: Attention to detail. You can start with checkpatch.pl, but do not
be mindlessly robotic in doing so. Re-read your own work as if you were the reviewer. If your change is a bug-fix, make sure your commit log indicates the end-user visible symptom, the underlying reason as to why it happens, and then if necessary, explain why the fix proposed
is the best way to get things done. Don't mangle whitespace, and as
is common, don't mis-indent function arguments that span multiple lines.
David Miller
maintainerFrom: Paul Gortmaker <[email protected]>
Date: Mon, 15 Jul 2013 22:59:54 -0400
Quoted 14 lines
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
Quoted 5 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Thanks a lot for writing this, it is useful.
I think it's easier if this lives somewhere under Documentation/networking in the kernel tree, then we can just point people at it.
What do you think?
Quoted 23 lines · David Miller
From: Paul Gortmaker <[email protected]>
Date: Mon, 15 Jul 2013 22:59:54 -0400Quoted 14 lines
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
Quoted 5 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Quoted 7 lines
Thanks a lot for writing this, it is useful.
I think it's easier if this lives somewhere under
Documentation/networkingin the kernel tree, then we can just point people at it.What do you think?
I was going to recommend the same. Ideally we would have sort of consistent documentation subsystem policy naming scheme to aid in finding and referencing documents.
Documentation/contributing.txtDocumentation/networking/contibuting.txtDocumentation/stable/contributing.txt
That might not be the best choice in naming, but consistency always makes things easier to find, especially in such a large system.
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel
Quoted 23 lines · David Miller
From: Paul Gortmaker <[email protected]>
Date: Mon, 15 Jul 2013 22:59:54 -0400Quoted 14 lines
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
Quoted 5 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Quoted 7 lines
Thanks a lot for writing this, it is useful.
I think it's easier if this lives somewhere under
Documentation/networkingin the kernel tree, then we can just point people at it.What do you think?
Maybe netdev_kernel_rules.txt, in keeping with the well-known stable_kernel_rules.txt?
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel
Quoted 23 lines · David Miller
From: Paul Gortmaker <[email protected]>
Date: Mon, 15 Jul 2013 22:59:54 -0400Quoted 14 lines
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
Quoted 5 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Quoted 7 lines
Thanks a lot for writing this, it is useful.
I think it's easier if this lives somewhere under
Documentation/networkingin the kernel tree, then we can just point people at it.What do you think?
Sounds fine to me, I'll make some of the corrections Darren pointed out and send it along as a patch.
Quoted 10 lines · Paul Gortmaker
Hi Dave,
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
Quoted 5 lines
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
And I'm going to learn how to properly interract on netdev while I'm at it. It's a win win :-)
Quoted 23 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Thanks,
Paul.
--Information you need to know about netdev
-----------------------------------------Q: What is netdev?
A: It is a mailing list for all network related linux stuff. This includes
anything found undernet/(i.e. core code like IPv6) anddrivers/net(i.e. hardware specific drivers) in the linux source tree.Note that some subsystems (e.g. wireless drivers) which have a high volume of traffic have their own specific mailing lists.
Quoted 9 lines
The netdev list is managed (like many other linux mailing lists) through VGER ( http://vger.kernel.org/ ) and archives can be found below:
http://marc.info/?l=linux-netdev
http://www.spinics.net/lists/netdev/Aside from subsystems like that mentioned above, all network related linux development (i.e. RFC, review, comments, etc) takes place on netdev.
Should LKML be Cc'd? I assume so...
Quoted 26 lines
Q: How do the changes posted to netdev make their way into linux?
A: There are always two trees (git repositories) in play. Both are driven
by David Miller, the main network maintainer. There is the "net" tree, and the "net-next" tree. As you can probably guess from the names, the net tree is for fixes to existing code already in the mainline tree from Linus, and net-next is where the new code goes for the future release. You can find the trees here:http://git.kernel.org/?p=linux/kernel/git/davem/net.git
http://git.kernel.org/?p=linux/kernel/git/davem/net-next.gitQ: How often do changes from these trees make it to the mainline Linus tree?
A: To understand this, you need to know a bit of background information
on the cadence of linux development. Each new release starts off with a two week "merge window" where the main maintainers feed their new stuff to Linus for merging into the mainline tree. After the two weeks, the merge window is closed, and it is called/tagged "-rc1". No new features get mainlined after this -- only fixes to the rc1 content are expected. After roughly a week of collecting fixes to the rc1 content, rc2 is released. This repeats on a roughly weekly basis until rc7 (typically; sometimes rc6 if things are quiet, or rc8 if things are in a state of churn), and a week after the last vX.Y-rcN was done, the official "vX.Y" is released.
Quoted 1 line
Relating that to netdev: At the beginning of the 2wk merge window,
Best not to abbreviate, we can spare the 3 bytes for " week" :-)
Quoted 11 lines
the net-next tree will be closed - no new changes/features. The accumulated new content of the past ~10 weeks will be passed onto mainline/Linus via a pull request for vX.Y -- at the same time, the "net" tree will start accumulating fixes for this pulled content relating to vX.Y
An announcement indicating when net-next has been closed is usually sent to netdev, but knowing the above, you can predict that in advance.
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
It would be handy to have a netdev-next bot that responded to "~/^ \[PATCH/" email (off list) during the merge window with a reminder of this point. Not everyone is active enough in kernel development be always aware of where we are in the cycle. Greg has a bot deal with common mistakes, so there is precedent.
Quoted 12 lines
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
The "net" tree continues to collect fixes for the vX.Y content, and is fed back to Linus at regular (~weekly) intervals. Meaning that the focus for "net" is on stablilization and bugfixes.
Finally, the vX.Y gets released, and the whole cycle starts over.
Q: So where are we now in this cycle?
Quoted 8 lines
A: Load the mainline (Linus) page here:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git
and note the top of the "tags" section. If it is rc1, it is early in the dev cycle. If it was tagged rc7 a week ago, then a release is probably imminent.
How does one determine if we are in the merge window? I wonder if we could have DEV_CYCLE file in the linux git repository which read:
MERGE WINDOW
BUG FIX ONLY (-rc1+)
CRITICAL FIXES ONLY (-rc4+)
That would make it trivial to know from right there in the sources where we are.
Quoted 43 lines
Q: How do I indicate which tree (net vs. net-next) my patch should be in?
A: Firstly, think whether you have a bug fix or new "next-like" content.
Then once decided, assuming that you use git, use the prefix flag, i.e.git format-patch --subject-prefix='PATCH net-next' start..finishUse "net" instead of "net-next" in the above for bug-fix net content. If you don't use git, then note the only magic in the above is just the subject text of the outgoing e-mail, and you can manually change it yourself with whatever MUA you are comfortable with.
Q: I sent a patch and I'm wondering what happened to it. How can I tell
whether it got merged?A: Start by looking at the main patchworks queue for netdev:
http://patchwork.ozlabs.org/project/netdev/list/
The "State" field will tell you exactly where things are at with your patch.
Q: The above only says "Under Review". How can I find out more?
A: Generally speaking, the patches get triaged quickly (in less than 48h).
So be patient. Asking the maintainer for status updates on your patch is a good way to ensure your patch is ignored or pushed to the bottom of the priority list.Q: How can I tell what patches are queued up for backporting to the
various stable releases?A: Normally Greg Kroah-Hartman collects stable commits himself, but
for networking, Dave collects up patches he deems critical for the networking subsystem, and then hands them off to Greg.There is a patchworks queue that you can see here:
http://patchwork.ozlabs.org/bundle/davem/stable/?state=*It contains the patches which Dave has selected, but not yet handed off to Greg. If Greg already has the patch, then it will be here:
http://git.kernel.org/cgit/linux/kernel/git/stable/stable-queue.git
Quoted 9 lines
A quick way to find whether the patch is in this stable-queue is to simply clone the repo, and then git grep the mainline commit ID, e.g. stable-queue$ git grep -l 284041ef21fdf2e releases/3.0.84/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.4.51/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.9.8/ipv6-fix-possible-crashes-in-ip6_cork_release.patch stable/stable-queue$
This needs a reference to stable_kernel_rules.txt IMO, and possibly less content here.
Quoted 3 lines
Q: I see a network patch and I think it should be backported to stable.
Should I request it via "[email protected]" like the references in the kernel'sDocumentation/directory say?
stable_kernel_rules.txt specifically
Quoted 4 lines
A: No, not for networking. Check the stable queues as per above 1st to see
if it is already queued. If not, then send a mail to netdev, listing the upstream commit ID and why you think it should be a stable candidate.
I had no idea as an infrequent contributor to netdev! stable_kernel_rules.txt needs some exceptions noted and a reference to this.
Quoted 22 lines
Before you jump to go do the above, do note that the normal stable rules in
Documentation/stable_kernel_rules.txtstill apply. So you need to explicitly indicate why it is a critical fix and exactly what users are impacted. In addition, you need to convince yourself that you _really_ think it has been overlooked, vs. having been considered and rejected.Generally speaking, the longer it has had a chance to "soak" in mainline, the better the odds that it is an OK candidate for stable. So scrambling to request a commit be added the day after it appears should be avoided.
Q: I have created a network patch and I think it should be backported to
stable. Should I add a "Cc: [email protected]" like the references in the kernel'sDocumentation/directory say?A: No. See above answer. In short, if you think it really belongs in
stable, then ensure you write a decent commit log that describes who gets impacted by the bugfix and how it manifests itself, and when the bug was introduced. If you do that properly, then the commit will get handled appropriately and most likely get put in the patchworks stable queue if it really warrants it.
Quoted 7 lines
If you think there is some valid information relating to it being in stable that does _not_ belong in the commit log, then use the three dash marker line as described in
Documentation/SubmittingPatchesto temporarily embed that information into the patch that you send.Q: Someone said that the comment style and coding convention is different
for the networing content. Is this true?
networking
Quoted 8 lines
A: Yes, in a largely trivial way. Instead of this:
/* * foobar blah blah blah * another line of text */
Quoted 5 lines
it is requested that you make it look like this:
/* foobar blah blah blah * another line of text */
This is.... unfortunate. I see the warnings from checkpatch.pl and I have to choose between adhering to that or keeping a file which is in complete violation to that consistent. I risk flaming either way.
Do we really need different coding styles for different sub directories of the same source tree? I won't say anything more on this or try to argue the point, it isn't my call. Just seems.... strange to me.
Quoted 6 lines
Q: I am working in existing code that has the former comment style and not the
latter. Should I submit new code in the former style or the latter?A: Make it the latter style, so that eventually all code in the domain of
netdev is of this format.
:-) OK
Quoted 11 lines
Q: I found a bug that might have possible security implications or similar.
Should I mail the main netdev maintainer off-list?A: No. The current netdev maintainer has consistently requested that people
use the mailing lists and not reach out directly. If you aren't OK with that, then perhaps consider using "[email protected]" instead.Q: What level of testing is expected before I submit my change?
A: If your changes are against net-next, then the expectation is that
s/then//
Quoted 7 lines
you have tested by layering your changes on top of net-next. Ideally you will have done run-time testing specific to your change, but at a minimum, your changes should survive an "allyesconfig" and an "allmodconfig" build without new warnings or failures.
Q: Any other tips to help ensure my
net/net-nextpatch gets OK'd?
Quoted 7 lines
A: Attention to detail. You can start with checkpatch.pl, but do not
be mindlessly robotic in doing so. Re-read your own work as if you were the reviewer. If your change is a bug-fix, make sure your commit log indicates the end-user visible symptom, the underlying reason as to why it happens, and then if necessary, explain why the fix proposed
is the best way to get things done. Don't mangle whitespace, and as
is common, don't mis-indent function arguments that span multiple lines.
This needs a reference to SubmittingPatches
This is great Paul, thank you for taking the time.
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel
Quoted 49 lines · Darren Hart
Quoted 10 lines · Paul Gortmaker
Hi Dave,
I was wondering if you think there is value in having a netdev-faq type document available -- perhaps as a vger mailout to new subscribers or similar? For example, I have lost count of the number of times that you have had to tell people that net-next is closed during the merge window. But you would probably be right in telling me that those same people don't read documentation. Well, that aside, I suppose answering my question is easier when there is a proposed starting point for content.
Quoted 5 lines
To that end, I've tried to collect a starting point based on repeated questions/corrections that I've seen over the years. I've added Greg to the Cc: in order to ensure I've captured the netdev-stable interaction correctly, and I've thrown Darren under the bus as a random content reviewer, since he has expressed an interest in documentation recently.
And I'm going to learn how to properly interract on netdev while I'm at it. It's a win win :-)
Quoted 29 lines
Below is a possible starting point for content. Many answers I have written are from memory, long after losing links to netdev threads that served as evidence for the answers, so apologies in advance if I have failed to recall correctly the specific details in which you would like to see things done.
Thanks,
Paul.
--Information you need to know about netdev
-----------------------------------------Q: What is netdev?
A: It is a mailing list for all network related linux stuff. This includes
anything found undernet/(i.e. core code like IPv6) anddrivers/net(i.e. hardware specific drivers) in the linux source tree.Note that some subsystems (e.g. wireless drivers) which have a high volume of traffic have their own specific mailing lists.
The netdev list is managed (like many other linux mailing lists) through VGER ( http://vger.kernel.org/ ) and archives can be found below:
http://marc.info/?l=linux-netdev
http://www.spinics.net/lists/netdev/
Quoted 6 lines
Quoted 3 lines
Aside from subsystems like that mentioned above, all network related linux development (i.e. RFC, review, comments, etc) takes place on netdev.
Should LKML be Cc'd? I assume so...
No, not unless there is a good reason to do so.
Quoted 28 lines
Quoted 28 lines
Q: How do the changes posted to netdev make their way into linux?
A: There are always two trees (git repositories) in play. Both are driven
by David Miller, the main network maintainer. There is the "net" tree, and the "net-next" tree. As you can probably guess from the names, the net tree is for fixes to existing code already in the mainline tree from Linus, and net-next is where the new code goes for the future release. You can find the trees here:http://git.kernel.org/?p=linux/kernel/git/davem/net.git
http://git.kernel.org/?p=linux/kernel/git/davem/net-next.gitQ: How often do changes from these trees make it to the mainline Linus tree?
A: To understand this, you need to know a bit of background information
on the cadence of linux development. Each new release starts off with a two week "merge window" where the main maintainers feed their new stuff to Linus for merging into the mainline tree. After the two weeks, the merge window is closed, and it is called/tagged "-rc1". No new features get mainlined after this -- only fixes to the rc1 content are expected. After roughly a week of collecting fixes to the rc1 content, rc2 is released. This repeats on a roughly weekly basis until rc7 (typically; sometimes rc6 if things are quiet, or rc8 if things are in a state of churn), and a week after the last vX.Y-rcN was done, the official "vX.Y" is released.
Quoted 4 lines
Quoted 1 line
Relating that to netdev: At the beginning of the 2wk merge window,
Best not to abbreviate, we can spare the 3 bytes for " week" :-)
Will fix.
Quoted 15 lines
Quoted 8 lines
the net-next tree will be closed - no new changes/features. The accumulated new content of the past ~10 weeks will be passed onto mainline/Linus via a pull request for vX.Y -- at the same time, the "net" tree will start accumulating fixes for this pulled content relating to vX.Y
Quoted 5 lines
An announcement indicating when net-next has been closed is usually sent to netdev, but knowing the above, you can predict that in advance.
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
Quoted 5 lines
It would be handy to have a netdev-next bot that responded to "~/^ \[PATCH/" email (off list) during the merge window with a reminder of this point. Not everyone is active enough in kernel development be always aware of where we are in the cycle. Greg has a bot deal with common mistakes, so there is precedent.
Perhaps, but that is outside of the scope of this document.
Quoted 27 lines
Quoted 14 lines
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
The "net" tree continues to collect fixes for the vX.Y content, and is fed back to Linus at regular (~weekly) intervals. Meaning that the focus for "net" is on stablilization and bugfixes.
Finally, the vX.Y gets released, and the whole cycle starts over.
Q: So where are we now in this cycle?
Quoted 8 lines
A: Load the mainline (Linus) page here:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git
and note the top of the "tags" section. If it is rc1, it is early in the dev cycle. If it was tagged rc7 a week ago, then a release is probably imminent.
How does one determine if we are in the merge window? I wonder if we could have
DEV_CYCLEfile in the linux git repository which read:
Quoted 6 lines
MERGE WINDOW
BUG FIX ONLY (-rc1+)
CRITICAL FIXES ONLY (-rc4+)That would make it trivial to know from right there in the sources where we are.
Again, outside the scope of this document.
Quoted 56 lines
Quoted 45 lines
Q: How do I indicate which tree (net vs. net-next) my patch should be in?
A: Firstly, think whether you have a bug fix or new "next-like" content.
Then once decided, assuming that you use git, use the prefix flag, i.e.git format-patch --subject-prefix='PATCH net-next' start..finishUse "net" instead of "net-next" in the above for bug-fix net content. If you don't use git, then note the only magic in the above is just the subject text of the outgoing e-mail, and you can manually change it yourself with whatever MUA you are comfortable with.
Q: I sent a patch and I'm wondering what happened to it. How can I tell
whether it got merged?A: Start by looking at the main patchworks queue for netdev:
http://patchwork.ozlabs.org/project/netdev/list/
The "State" field will tell you exactly where things are at with your patch.
Q: The above only says "Under Review". How can I find out more?
A: Generally speaking, the patches get triaged quickly (in less than 48h).
So be patient. Asking the maintainer for status updates on your patch is a good way to ensure your patch is ignored or pushed to the bottom of the priority list.Q: How can I tell what patches are queued up for backporting to the
various stable releases?A: Normally Greg Kroah-Hartman collects stable commits himself, but
for networking, Dave collects up patches he deems critical for the networking subsystem, and then hands them off to Greg.There is a patchworks queue that you can see here:
http://patchwork.ozlabs.org/bundle/davem/stable/?state=*It contains the patches which Dave has selected, but not yet handed off to Greg. If Greg already has the patch, then it will be here:
http://git.kernel.org/cgit/linux/kernel/git/stable/stable-queue.gitQuoted 9 lines
A quick way to find whether the patch is in this stable-queue is to simply clone the repo, and then git grep the mainline commit ID, e.g. stable-queue$ git grep -l 284041ef21fdf2e releases/3.0.84/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.4.51/ipv6-fix-possible-crashes-in-ip6_cork_release.patch releases/3.9.8/ipv6-fix-possible-crashes-in-ip6_cork_release.patch stable/stable-queue$
Quoted 2 lines
This needs a reference to stable_kernel_rules.txt IMO, and possibly less content here.
The question is about finding whether a patch is queued, which comes up quite often, so I think the detail is warranted. The rules file is more about requirements for getting a patch _in_ stable, so the reference addition below makes sense.
Quoted 7 lines
Quoted 5 lines
Q: I see a network patch and I think it should be backported to stable.
Should I request it via "[email protected]" like the references in the kernel'sDocumentation/directory say?stable_kernel_rules.txt specifically
Will add that.
Quoted 10 lines
Quoted 5 lines
A: No, not for networking. Check the stable queues as per above 1st to see
if it is already queued. If not, then send a mail to netdev, listing the upstream commit ID and why you think it should be a stable candidate.I had no idea as an infrequent contributor to netdev! stable_kernel_rules.txt needs some exceptions noted and a reference to this.
Send Greg a patch. :)
Quoted 29 lines
Quoted 29 lines
Before you jump to go do the above, do note that the normal stable rules in
Documentation/stable_kernel_rules.txtstill apply. So you need to explicitly indicate why it is a critical fix and exactly what users are impacted. In addition, you need to convince yourself that you _really_ think it has been overlooked, vs. having been considered and rejected.Generally speaking, the longer it has had a chance to "soak" in mainline, the better the odds that it is an OK candidate for stable. So scrambling to request a commit be added the day after it appears should be avoided.
Q: I have created a network patch and I think it should be backported to
stable. Should I add a "Cc: [email protected]" like the references in the kernel'sDocumentation/directory say?A: No. See above answer. In short, if you think it really belongs in
stable, then ensure you write a decent commit log that describes who gets impacted by the bugfix and how it manifests itself, and when the bug was introduced. If you do that properly, then the commit will get handled appropriately and most likely get put in the patchworks stable queue if it really warrants it.If you think there is some valid information relating to it being in stable that does _not_ belong in the commit log, then use the three dash marker line as described in
Documentation/SubmittingPatchesto temporarily embed that information into the patch that you send.
Quoted 5 lines
Quoted 2 lines
Q: Someone said that the comment style and coding convention is different
for the networing content. Is this true?networking
fixed.
Quoted 16 lines
Quoted 9 lines
A: Yes, in a largely trivial way. Instead of this:
/* * foobar blah blah blah * another line of text */Quoted 5 lines
it is requested that you make it look like this:
/* foobar blah blah blah * another line of text */
Quoted 7 lines
This is.... unfortunate. I see the warnings from checkpatch.pl and I have to choose between adhering to that or keeping a file which is in complete violation to that consistent. I risk flaming either way.
Do we really need different coding styles for different sub directories of the same source tree? I won't say anything more on this or try to argue the point, it isn't my call. Just seems.... strange to me.
It is what it is; I'm just documenting it here.
Quoted 21 lines
Quoted 8 lines
Q: I am working in existing code that has the former comment style and not the
latter. Should I submit new code in the former style or the latter?A: Make it the latter style, so that eventually all code in the domain of
netdev is of this format.:-) OK
Quoted 8 lines
Q: I found a bug that might have possible security implications or similar.
Should I mail the main netdev maintainer off-list?A: No. The current netdev maintainer has consistently requested that people
use the mailing lists and not reach out directly. If you aren't OK with that, then perhaps consider using "[email protected]" instead.
Quoted 5 lines
Quoted 3 lines
Q: What level of testing is expected before I submit my change?
A: If your changes are against net-next, then the expectation is that
s/then//
Fixed.
Quoted 17 lines
Quoted 8 lines
you have tested by layering your changes on top of net-next. Ideally you will have done run-time testing specific to your change, but at a minimum, your changes should survive an "allyesconfig" and an "allmodconfig" build without new warnings or failures.
Q: Any other tips to help ensure my
net/net-nextpatch gets OK'd?Quoted 7 lines
A: Attention to detail. You can start with checkpatch.pl, but do not
be mindlessly robotic in doing so. Re-read your own work as if you were the reviewer. If your change is a bug-fix, make sure your commit log indicates the end-user visible symptom, the underlying reason as to why it happens, and then if necessary, explain why the fix proposed
is the best way to get things done. Don't mangle whitespace, and as
is common, don't mis-indent function arguments that span multiple lines.
Quoted 1 line
This needs a reference to SubmittingPatches
Sure, that can't hurt either. Will add it.
Paul.
--
Quoted 3 lines
This is great Paul, thank you for taking the time.
Quoted 5 lines · Paul Gortmaker
Quoted 3 lines
Do we really need different coding styles for different sub directories of the same source tree? I won't say anything more on this or try to argue the point, it isn't my call. Just seems.... strange to me.
It is what it is; I'm just documenting it here.
Not that duplication is bad, but that comment style variation is documented in CodingStyle.
Quoted 11 lines · Paul Gortmaker
Quoted 7 lines · Darren Hart
Quoted 4 lines
Aside from subsystems like that mentioned above, all network related linux development (i.e. RFC, review, comments, etc) takes place on netdev.
Should LKML be Cc'd? I assume so...
No, not unless there is a good reason to do so.
That advice would run counter to what's inDocumentation/SubmittingPatches:
6) Select your CC (e-mail carbon copy) list.
Unless you have a reason NOT to do so, CC [email protected].
Also, get-maintainers includes lkml all the time.
Quoted 4 lines · Paul Gortmaker
Q: Any other tips to help ensure my
net/net-nextpatch gets OK'd?A: Attention to detail. You can start with checkpatch.pl, but do not
be mindlessly robotic in doing so.
Suggest checkpatch.pl --strict because there are some style elements (like multi-line statement alignment) where checkpatch bleats with --strict that are commonly used in net and drivers/net.
email the proposed patch to yourself and make sure it applies cleanly before sending it to the netdev mailing list.
Quoted 5 lines · Paul Gortmaker
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
So I misunderstood this line apparently. I checked Linus tree:
$ git show v3.11-rc1 tag v3.11-rc1 Tagger: Linus Torvalds <[email protected]> Date: Sun Jul 14 15:18:41 2013 -0700
Linux 3.11-rc1
And seeing that rc1 had been tagged 3 days ago, submitted the pch_gbe patch for MinnowBoard to netdev for net-next. David *very* politely :-) asked me to resubmit when the net-next tree was open. It appears to me then that we need to add a bit here about searching the archives mentioned above for the last netdev-is-open message (I presume there is a common subject format) to know whether a patch will be welcomed or not.
I am not subscribed to netdev as I contribute only very rarely. Is it considered a requirement that one be subscribed to netdev in order to contribute? If so, we should add that here as well.
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel
Quoted 17 lines · Darren Hart
Quoted 5 lines · Paul Gortmaker
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
So I misunderstood this line apparently. I checked Linus tree:
$ git show v3.11-rc1 tag v3.11-rc1 Tagger: Linus Torvalds <[email protected]> Date: Sun Jul 14 15:18:41 2013 -0700Linux 3.11-rc1
Quoted 7 lines
And seeing that rc1 had been tagged 3 days ago, submitted the
pch_gbepatch for MinnowBoard to netdev for net-next. David *very* politely :-) asked me to resubmit when the net-next tree was open. It appears to me then that we need to add a bit here about searching the archives mentioned above for the last netdev-is-open message (I presume there is a common subject format) to know whether a patch will be welcomed or not.
I will append under the "Shortly after..." text something like:
"If you aren't subscribed to netdev and/or are simply unsure if net-next has re-opened yet, simply check the net-next git repository link above for any new networking related commits."
That "link above" would be this:
http://git.kernel.org/cgit/linux/kernel/git/davem/net-next.git
P.
--
Quoted 6 lines
I am not subscribed to netdev as I contribute only very rarely. Is it considered a requirement that one be subscribed to netdev in order to contribute? If so, we should add that here as well.
Quoted 8 lines · Darren Hart
Quoted 5 lines · Paul Gortmaker
IMPORTANT: Do not send new net-next content to netdev during the period during which net-next tree is closed.
Shortly after the two weeks have passed, (and vX.Y-rc1 is released) the tree for net-next reopens to collect content for the next (vX.Y+1) release.
yesterday I tried to send an email to Jeremy Kerr asking if there was some way to tweak patchwork to allow DaveM a way to signal net-next being closed - perhaps some message at http://patchwork.ozlabs.org/project/netdev/list/ Baring that, I wondered if it would be possible to submit a dummy "[PATCH net-next] net-next closed until futher notice" patch, and fake-out the date to be some point in the future, so it would appear at the top of the list of patches.
rick jones
Thanks for the great FAQ!
Quoted 6 lines · Paul Gortmaker
Q: I found a bug that might have possible security implications or similar.
Should I mail the main netdev maintainer off-list?A: No. The current netdev maintainer has consistently requested that people
use the mailing lists and not reach out directly. If you aren't OK with that, then perhaps consider using "[email protected]" instead.
I don't have a strong opinion on that, just a proposal (and I certainly don't want to start another discussion about that):
Maybe we could give a hint about the linux-distros mailing list here, too?
<http://oss-security.openwall.org/wiki/mailing-lists/distros>
Thanks,
Hannes