Request for reviews for 4.4-st55, 4.4-cip90

Hi!

Here's the manual backports to review for the next 4.4 release which, as always, can be found in https://git.kernel.org/pub/scm/linux/kernel/git/cip/linux-cip.git/log/?h=linux-4.4.y-st-rc

f419702e01650 media: mc: mark the media devnode as registered from the, start 65ea5a47714a3 nilfs2: fix nilfs_empty_dir() misjudgment and long loop on I/O errors dff699deb3395 HID: core: remove unnecessary WARN_ON() in implement() 5e5344dac2f00 usb: misc: uss720: check for incompatible versions of the Belkin F5U002 f13500b98a507 udf: udftime: prevent overflow in udf_disk_stamp_to_time() 2efb5ca24b08f net: usb: rtl8150 fix unintiatilzed variables in rtl8150_get_link_ksettings 3ffc17009d168 ARM: dts: samsung: smdkv310: fix keypad no-autorepeat cccb925ba214a ARM: dts: samsung: exynos4412-origen: fix keypad no-autorepeat a666ac1afa8ca ARM: dts: samsung: smdk4412: fix keypad no-autorepeat a950059ab72ae netfilter: nf_tables: fully validate NFT_DATA_VALUE on store to data registers 9a626ec9815e8 batman-adv: Don't accept TT entries for out-of-spec VIDs 18df69aa504b6 ftruncate: pass a signed offset
0ca4b05662224 ipv6: annotate some data-races around sk->sk_prot
b04d56f62c5dd ipv6: Fix data races around sk->sk_prot.
9193eb58d4dc0 UPSTREAM: tcp: fix DSACK undo in fast recovery to call tcp_try_to_open() d467eda9f8982 nilfs2: fix inode number range checks
0c5c578a19ef6 i2c: pnx: Fix potential deadlock warning from del_timer_sync() call in isr 8267db6d29ea8 nilfs2: fix incorrect inode allocation from reserved inodes 1df198bc10b67 libceph: fix race between delayed_work() and ceph_monc_stop() 66d3371a26e02 SUNRPC: Fix RPC client cleaned up the freed pipefs dentries e2be94637020f fs/file: fix the check in find_next_fd()
d469b14c3f2ad fs: better handle deep ancestor chains in is_subdir()

There is also a fix-up for an error in a previous backport:

3ffe90bcdadb8 nilfs2: fix incorrect use of nilfs_error()

Thanks for having a look.

CU
Uli

Hi!

Thanks for series.

Most of these look good, but I believe we should just drop "5e5344dac2f00 usb: misc: uss720: check for incompatible versions of the Belkin F5U002". First, it simply drops support for some hardware without really understanding the problem. Second, AFAICT error handling is wrong, there's probe_abort label that does some kind of cleanup but this just returns directly. Problem likely exists in mainline, too. Should I try to report it?

Best regards,
Pavel

a just an API fix > f419702e01650 media: mc: mark the media devnode as registered from the, start
a > 65ea5a47714a3 nilfs2: fix nilfs_empty_dir() misjudgment and long loop on I/O errors
a just a WARN removal > dff699deb3395 HID: core: remove unnecessary WARN_ON() in implement()
!! this is not right; should goto probe abort; this may break someone's config > 5e5344dac2f00 usb: misc: uss720: check for incompatible versions of the Belkin F5U002 a just a robustness> f13500b98a507 udf: udftime: prevent overflow in udf_disk_stamp_to_time()
a > 2efb5ca24b08f net: usb: rtl8150 fix unintiatilzed variables in rtl8150_get_link_ksettings
a just a autorepeat tweak > 3ffc17009d168 ARM: dts: samsung: smdkv310: fix keypad no-autorepeat
a just a autorepeat tweak > cccb925ba214a ARM: dts: samsung: exynos4412-origen: fix keypad no-autorepeat
a just a autorepeat tweak > a666ac1afa8ca ARM: dts: samsung: smdk4412: fix keypad no-autorepeat
a > a950059ab72ae netfilter: nf_tables: fully validate NFT_DATA_VALUE on store to data registers
a > 9a626ec9815e8 batman-adv: Don't accept TT entries for out-of-spec VIDs
a just an API fix > 18df69aa504b6 ftruncate: pass a signed offset
a just a _ONCE annotation > 0ca4b05662224 ipv6: annotate some data-races around sk->sk_prot
a just a _ONCE annotation > b04d56f62c5dd ipv6: Fix data races around sk->sk_prot.
a > 9193eb58d4dc0 UPSTREAM: tcp: fix DSACK undo in fast recovery to call tcp_try_to_open()
a > d467eda9f8982 nilfs2: fix inode number range checks
a > 0c5c578a19ef6 i2c: pnx: Fix potential deadlock warning from del_timer_sync() call in isr
a > 8267db6d29ea8 nilfs2: fix incorrect inode allocation from reserved inodes
a not a minimum fix> 1df198bc10b67 libceph: fix race between delayed_work() and ceph_monc_stop()
a > 66d3371a26e02 SUNRPC: Fix RPC client cleaned up the freed pipefs dentries
a just a tiny performance tweak > e2be94637020f fs/file: fix the check in find_next_fd()
a > d469b14c3f2ad fs: better handle deep ancestor chains in is_subdir()

a > 3ffe90bcdadb8 nilfs2: fix incorrect use of nilfs_error()

DENX Software Engineering GmbH,        Managing Director: Erika Unter
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany

Thank you for the review.

It looks more like the commit message and the comment are misleading.
The patch doesn't actually special-case a specific incompatible device, it just bails out if get_1284_register() fails.

Yes, that actually looks like a bug.

I think so.

CU
Uli