Skip to contentSkip to conversation

linux-next: manual merge of the mtd tree with the vfs-brauner tree

DISCUSSION Stephen Rothwell 5 participants 14 messages Oct 29, 2023 – Oct 31, 2023 Linux-next tree

lore
Share & tools
lore
#0 lore raw

Hi all,

Today's linux-next merge of the mtd tree got a conflict in:

drivers/mtd/devices/block2mtd.c

between commit:

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

from the mtd tree.

I fixed it up (I just used the former) and can carry the fix as necessary. This is now fixed as far as linux-next is concerned, but any non trivial conflicts should be mentioned to your upstream maintainer when your tree is submitted for merging. You may also want to consider cooperating with the maintainer of the conflicting tree to minimise any particularly complex conflicts.

Signature · 2 lines
Cheers,
Stephen Rothwell
#1 lore raw

Hi,

[email protected] wrote on Mon, 30 Oct 2023 10:34:15 +1100:

Quoted 9 lines

Hi all,

Today's linux-next merge of the mtd tree got a conflict in:

drivers/mtd/devices/block2mtd.c

between commit:

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Quoted 3 lines

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

I will drop this commit from mtd/next. Please take it through the vfs-brauner tree as well to avoid conflicts or otherwise, Richard, can you send an update at -rc1?

Thanks,
Miquèl

#2 lore raw

----- Ursprüngliche Mail -----

Quoted 10 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 7 lines

Today's linux-next merge of the mtd tree got a conflict in:

drivers/mtd/devices/block2mtd.c

between commit:

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Me neither. :-/

Quoted 7 lines
Quoted 3 lines

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

I will drop this commit from mtd/next. Please take it through the vfs-brauner tree as well to avoid conflicts or otherwise, Richard, can you send an update at -rc1?

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()") is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Thanks,
//richard

#3 lore raw
Quoted 6 lines · Richard Weinberger

----- Ursprüngliche Mail -----

Quoted 5 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 4 lines

Today's linux-next merge of the mtd tree got a conflict in:

drivers/mtd/devices/block2mtd.c

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

between commit:

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Me neither. :-/

I'm sorry for that but I took the maintainers entry for BLOCK2MTD which is:

BLOCK2MTD DRIVER

M:      Joern Engel <[email protected]>
L:      [email protected]
S:      Maintained
F:      drivers/mtd/devices/block2mtd.c

And both Joern and linux-mtd were CCed on the patch. If different people should be CCed these days, please update the entry. Thanks!

Quoted 8 lines
Quoted 7 lines
Quoted 3 lines

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

I will drop this commit from mtd/next. Please take it through the vfs-brauner tree as well to avoid conflicts or otherwise, Richard, can you send an update at -rc1?

Quoted 7 lines

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()") is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Yes, that's one of the cases where stable rules make life harder for actual fixes... You can try pushing ff6abbe85634 to stable even if it is not upstream since it fixes a real bug and taking the upstream solution is indeed IMO too intrusive. Sometimes stable maintainers accept such fixes.

Honza

Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR
#4 lore raw

Jan,

----- Ursprüngliche Mail -----

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Me neither. :-/

I'm sorry for that but I took the maintainers entry for BLOCK2MTD which is:

Quoted 8 lines

BLOCK2MTD DRIVER

M:      Joern Engel <[email protected]>
L:      [email protected]
S:      Maintained
F:      drivers/mtd/devices/block2mtd.c

And both Joern and linux-mtd were CCed on the patch. If different people should be CCed these days, please update the entry. Thanks!

Ah, you did a manual lookup?
Because get_maintainer.pl seems to do the right thing:

$ ./scripts/get_maintainer.pl drivers/mtd/devices/block2mtd.c
Joern Engel <[email protected]> (maintainer:BLOCK2MTD DRIVER)
Miquel Raynal <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Richard Weinberger <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Vignesh Raghavendra <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
[email protected] (open list:BLOCK2MTD DRIVER)
[email protected] (open list)
Quoted 17 lines
Quoted 8 lines
Quoted 7 lines
Quoted 3 lines

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

I will drop this commit from mtd/next. Please take it through the vfs-brauner tree as well to avoid conflicts or otherwise, Richard, can you send an update at -rc1?

Quoted 8 lines

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")
is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Quoted 4 lines

Yes, that's one of the cases where stable rules make life harder for actual fixes... You can try pushing ff6abbe85634 to stable even if it is not upstream since it fixes a real bug and taking the upstream solution is indeed IMO too intrusive. Sometimes stable maintainers accept such fixes.

Yep, let's try this route. :-)

Thanks,
//richard

#5 lore raw
Quoted 48 lines · Richard Weinberger

Jan,

----- Ursprüngliche Mail -----

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Me neither. :-/

I'm sorry for that but I took the maintainers entry for BLOCK2MTD which is:

Quoted 8 lines

BLOCK2MTD DRIVER

M:      Joern Engel <[email protected]>
L:      [email protected]
S:      Maintained
F:      drivers/mtd/devices/block2mtd.c

And both Joern and linux-mtd were CCed on the patch. If different people should be CCed these days, please update the entry. Thanks!

Ah, you did a manual lookup?
Because get_maintainer.pl seems to do the right thing:

$ ./scripts/get_maintainer.pl drivers/mtd/devices/block2mtd.c
Joern Engel <[email protected]> (maintainer:BLOCK2MTD DRIVER)
Miquel Raynal <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Richard Weinberger <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Vignesh Raghavendra <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
[email protected] (open list:BLOCK2MTD DRIVER)
[email protected] (open list)
Quoted 17 lines
Quoted 16 lines
Quoted 7 lines
Quoted 3 lines

from the vfs-brauner tree and commit:

ff6abbe85634 ("mtd: block2mtd: Add a valid holder to blkdev_put()")

I will drop this commit from mtd/next. Please take it through the vfs-brauner tree as well to avoid conflicts or otherwise, Richard, can you send an update at -rc1?

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")
is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Quoted 6 lines
Quoted 4 lines

Yes, that's one of the cases where stable rules make life harder for actual fixes... You can try pushing ff6abbe85634 to stable even if it is not upstream since it fixes a real bug and taking the upstream solution is indeed IMO too intrusive. Sometimes stable maintainers accept such fixes.

Yep, let's try this route. :-)

Is there anything for me to do? IOW, do I need to grab that patch or not? :)

#6 lore raw

Christian,

----- Ursprüngliche Mail -----

Quoted 15 lines

Von: "Christian Brauner" <[email protected]>

Quoted 14 lines
Quoted 13 lines
Quoted 8 lines

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")
is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Yes, that's one of the cases where stable rules make life harder for actual fixes... You can try pushing ff6abbe85634 to stable even if it is not upstream since it fixes a real bug and taking the upstream solution is indeed IMO too intrusive. Sometimes stable maintainers accept such fixes.

Quoted 4 lines
Quoted 1 line

Yep, let's try this route. :-)

Is there anything for me to do? IOW, do I need to grab that patch or not? :)

No, just keep Jan's patch. (-:

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

Thanks,
//richard

#7 lore raw

Hi Richard,

[email protected] wrote on Tue, 31 Oct 2023 12:30:40 +0100 (CET):

Quoted 20 lines

Christian,

----- Ursprüngliche Mail -----

Quoted 17 lines

Von: "Christian Brauner" <[email protected]>

Quoted 15 lines
Quoted 9 lines
Quoted 8 lines

A side effect of 1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")
is that it fixes the problem too. That's a good thing.

I'm a bit puzzled how to fix the problem for 6.5.y and 6.6.y stable releases. Back porting 1bcded92d938 seems risky to me since the commit is large. On the other hand, ff6abbe85634 will not make it into Linus' tree and therefore is not suitable for stable either.

Quoted 4 lines

Yes, that's one of the cases where stable rules make life harder for actual fixes... You can try pushing ff6abbe85634 to stable even if it is not upstream since it fixes a real bug and taking the upstream solution is indeed IMO too intrusive. Sometimes stable maintainers accept such fixes.

Yep, let's try this route. :-)

Quoted 7 lines
Quoted 2 lines

Is there anything for me to do? IOW, do I need to grab that patch or not? :)

No, just keep Jan's patch. (-:

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

It's not relevant if the patch in Brauner's tree is already fixing this up. Just send the smaller patch to [email protected] asking them to backport this patch instead of the other one, they are used to this kind of constraint, no?

Miquèl

#8 lore raw

----- Ursprüngliche Mail -----

Quoted 8 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 2 lines

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

It's not relevant if the patch in Brauner's tree is already fixing this up. Just send the smaller patch to [email protected] asking them to backport this patch instead of the other one, they are used to this kind of constraint, no?

I'm just in fear of stable rule #1.
"It or an equivalent fix must already exist in Linus' tree (upstream)."

Anyway, I'll try.

Thanks,
//richard

#9 lore raw

[email protected] wrote on Tue, 31 Oct 2023 14:13:53 +0100 (CET):

Quoted 5 lines

----- Ursprüngliche Mail -----

Quoted 4 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 2 lines

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

Quoted 7 lines
Quoted 4 lines

It's not relevant if the patch in Brauner's tree is already fixing this up. Just send the smaller patch to [email protected] asking them to backport this patch instead of the other one, they are used to this kind of constraint, no?

I'm just in fear of stable rule #1.
"It or an equivalent fix must already exist in Linus' tree (upstream)."

It should be very soon, the merge window is open ;)

Cheers,
Miquèl

#10 lore raw
Quoted 13 lines · Miquel Raynal

[email protected] wrote on Tue, 31 Oct 2023 14:13:53 +0100 (CET):

Quoted 10 lines

----- Ursprüngliche Mail -----

Quoted 8 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 2 lines

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

It's not relevant if the patch in Brauner's tree is already fixing this up. Just send the smaller patch to [email protected] asking them to backport this patch instead of the other one, they are used to this kind of constraint, no?

Quoted 4 lines
Quoted 2 lines

I'm just in fear of stable rule #1.
"It or an equivalent fix must already exist in Linus' tree (upstream)."

It should be very soon, the merge window is open ;)

vfs-6.7.super was merged yesterday, if that's what this is about.

#11 lore raw

Hi Christian,

[email protected] wrote on Tue, 31 Oct 2023 15:25:19 +0100:

Quoted 14 lines · Miquel Raynal
Quoted 13 lines

[email protected] wrote on Tue, 31 Oct 2023 14:13:53 +0100 (CET):

Quoted 10 lines

----- Ursprüngliche Mail -----

Quoted 8 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 2 lines

Miquel, we could also keep ff6abbe85634 in the mtd tree and explain Linus the conflict, what do you think? That would help with back porting to stable.

It's not relevant if the patch in Brauner's tree is already fixing this up. Just send the smaller patch to [email protected] asking them to backport this patch instead of the other one, they are used to this kind of constraint, no?

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

I'm just in fear of stable rule #1.
"It or an equivalent fix must already exist in Linus' tree (upstream)."

It should be very soon, the merge window is open ;)

vfs-6.7.super was merged yesterday, if that's what this is about.

Great, thanks for the confirmation. Then Richard you have your upstream commit already!

Thanks,
Miquèl

#12 lore raw

----- Ursprüngliche Mail -----

Quoted 5 lines

Von: "Miquel Raynal" <[email protected]>

Quoted 1 line

vfs-6.7.super was merged yesterday, if that's what this is about.

Great, thanks for the confirmation. Then Richard you have your upstream commit already!

Well, I'd like to back port ff6abbe85634, which is not upstream.
But I'll try to argue that ff6abbe85634 is the super simple variant of 1bcded92d938.

Thanks,
//richard

#13 lore raw
In reply to Richard Weinberger
Quoted 17 lines · Richard Weinberger

Jan,

----- Ursprüngliche Mail -----

Quoted 14 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

1bcded92d938 ("mtd: block2mtd: Convert to bdev_open_by_dev/path()")

I haven't seen this commit, I was not Cc'ed.

Me neither. :-/

I'm sorry for that but I took the maintainers entry for BLOCK2MTD which is:

BLOCK2MTD DRIVER

M:      Joern Engel <[email protected]>
L:      [email protected]
S:      Maintained
F:      drivers/mtd/devices/block2mtd.c
Quoted 4 lines
Quoted 2 lines

And both Joern and linux-mtd were CCed on the patch. If different people should be CCed these days, please update the entry. Thanks!

Ah, you did a manual lookup?

Yes.

Quoted 9 lines

Because get_maintainer.pl seems to do the right thing:

$ ./scripts/get_maintainer.pl drivers/mtd/devices/block2mtd.c
Joern Engel <[email protected]> (maintainer:BLOCK2MTD DRIVER)
Miquel Raynal <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Richard Weinberger <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
Vignesh Raghavendra <[email protected]> (maintainer:MEMORY TECHNOLOGY DEVICES (MTD))
[email protected] (open list:BLOCK2MTD DRIVER)
[email protected] (open list)

Hum, right. I didn't realize there could be multiple entries matching one file. My mistake.

Honza

Signature · 2 lines
Jan Kara <[email protected]>
SUSE Labs, CR