Now that d-i rc1 is out, it's time to update the kernel in unstable.
There are a huge number of pending fixes from 3.2.y (including security
issues and the Samsung laptop brick bug), lots of new hardware support
(particularly the DRM update), and udeb updates for armhf and for HID
drivers. The curent changelog entry is below.
I expect to release 3.2.39 on Tuesday, having started the review cycle
this evening. This should include fixes for CVE-2013-0216,
CVE-2013-0217, CVE-2013-0228 and CVE-2013-0871.
Would it be OK to upload a package based on that shortly after? Are
there any other urgent fixes?
Ben.
---
linux (3.2.38-1) UNRELEASED; urgency=low
* New upstream stable update:
http://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.2.36
- freezer: PF_FREEZER_NOSIG should be cleared along with PF_NOFREEZE
(Closes: #697077)
- tmpfs: fix shared mempolicy leak
- virtio: 9p: correctly pass physical address to userspace for high pages
- virtio: force vring descriptors to be allocated from lowmem
- USB: EHCI: bugfix: urb->hcpriv should not be NULL
- rcu: Fix batch-limit size problem
- Bluetooth: ath3k: Add support for VAIO VPCEH [0489:e027]
(Closes: #700550)
- mvsas: fix undefined bit shift
- ALSA: usb-audio: Avoid autopm calls after disconnection; Fix missing
autopm for MIDI input (Closes: #664068)
- target/file: Fix 32-bit highmem breakage for SGL -> iovec mapping
- SCSI: fix Null pointer dereference on disk error
- proc: pid/status: show all supplementary groups
- nfsd4: fix oops on unusual readlike compound
- ARM: missing ->mmap_sem around find_vma() in swp_emulate.c
- sctp: fix memory leak in sctp_datamsg_from_user() when copy from user
space fails
- ne2000: add the right platform device
- irda: sir_dev: Fix copy/paste typo
- ipv4: ip_check_defrag must not modify skb before unsharing
- telephony: ijx: buffer overflow in ixj_write_cid()
- udf: fix memory leak while allocating blocks during write
http://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.2.37
- ext4: fix extent tree corruption caused by hole punch
- jbd2: fix assertion failure in jbd2_journal_flush()
- tmpfs mempolicy: fix /proc/mounts corrupting memory
- sparc: huge_ptep_set_* functions need to call set_huge_pte_at()
- inet: Fix kmemleak in tcp_v4/6_syn_recv_sock and
dccp_v4/6_request_recv_sock
- net: sched: integer overflow fix
- tcp: implement RFC 5961 3.2
- tcp: implement RFC 5961 4.2
- tcp: refine SYN handling in tcp_validate_incoming
- tcp: tcp_replace_ts_recent() should not be called from
tcp_validate_incoming()
- tcp: RFC 5961 5.2 Blind Data Injection Attack Mitigation
- RDMA/nes: Fix for crash when registering zero length MR for CQ
- ACPI : do not use Lid and Sleep button for S5 wakeup
http://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.2.38
- staging: comedi: comedi_test: fix race when cancelling command
- mm: use aligned zone start for pfn_to_bitidx calculation
- [s390] s390/time: fix sched_clock() overflow (Closes: #698382)
- [i386] xen: Fix stack corruption in xen_failsafe_callback for 32bit
PVOPS guests. (CVE-2013-0190)
- KVM: PPC: Emulate dcbf
- evm: checking if removexattr is not a NULL
- ath9k_htc: Fix memory leak
- ath9k: do not link receive buffers during flush
- ath9k: fix double-free bug on beacon generate failure
- x86/msr: Add capabilities check
- can: c_can: fix invalid error codes
- can: ti_hecc: fix invalid error codes
- can: pch_can: fix invalid error codes
- smp: Fix SMP function call empty cpu mask race
- xfs: Fix possible use-after-free with AIO
- EDAC: Test correct variable in ->store function
- samsung-laptop: Disable on EFI hardware, to avoid damaging it
- NFS: Don't silently fail setattr() requests on mountpoints
- intel-iommu: Prevent devices with RMRRs from being placed into SI Domain
- ALSA: usb-audio: Fix regression by disconnection-race-fix patch
(Closes: #696321)
- printk: fix buffer overflow when calling log_prefix function from
call_console_drivers
[ Ben Hutchings ]
* Input: wacom - fix touch support for Bamboo Fun CTH-461
* media/rc: Add iguanair driver from Linux 3.7 (Closes: #696925)
* rt2800: add chipset revision RT5390R support (Closes: #696592)
* [armhf/mx5] mtd: Enable MTD_BLOCK as module
* [armhf/mx5] udeb: Add missing storage drivers (Closes: #697128)
- Add ata-modules including libata, pata-modules including pata_imx,
sata-modules including ahci_platform
- Add sdhci-esdhc-imx to mmc-modules
- Add mtd-modules including mtd, mtdblock and m25p80
* [armhf] udeb: Fix network driver selection
- [armhf/mx5] Remove nic-modules
- [armhf/vexpress] Add usb-modules
- Add standard set of USB drivers to nic-usb-modules
- Add nic-wireless-modules
* be2net: Apply backported fixes requested by Emulex (Closes: #697479)
- be2net: do not modify PCI MaxReadReq size
- be2net: fix reporting number of actual rx queues
- be2net: do not use SCRATCHPAD register
- be2net: reduce gso_max_size setting to account for ethernet header.
- be2net: Increase statistics structure size for skyhawk.
- be2net: Explicitly clear the reserved field in the Tx Descriptor
- be2net: Regression bug wherein VFs creation broken for multiple cards.
- be2net: Fix to trim skb for padded vlan packets to workaround an ASIC Bug
- be2net: Fix Endian
- be2net: Enable RSS UDP hashing for Lancer and Skyhawk
- be2net: dont pull too much data in skb linear part
- be2net: Fix to parse RSS hash from Receive completions correctly.
- be2net: Avoid disabling BH in be_poll()
* udeb: Add specialised USB keyboard/mouse drivers to input-modules:
hid-a4tech, hid-cypress, hid-ezkey (Closes: #697035), hid-kensington,
hid-keytouch, hid-kye, hid-multitouch, hid-ortek, hid-primax,
hid-quanta, hid-samsung, hid-speedlink
* radeon: Firmware is required for DRM and KMS on R600 onward, but not
for KMS on earlier chips (Closes: #697229)
* [!powerpc] radeon: Reenable DRM_RADEON_KMS, as it apparently works on
most non-PowerMac systems
* fs: cachefiles: add support for large files in filesystem caching
(Closes: #698376)
* [rt] Update to 3.2.38-rt57:
- sched: Adjust sched_reset_on_fork when nothing else changes
- sched: Queue RT tasks to head when prio drops
- sched: Consider pi boosting in setscheduler
- sched: Init idle->on_rq in init_idle()
- sched: Check for idle task in might_sleep()
- mm: swap: Initialize local locks early
* bridge: Pull ip header into skb->data before looking into ip header.
(Closes: #697903)
* [armel/versatile,armhf/vexpress] i2c: Enable I2C, I2C_VERSATILE as modules
(Closes: #696182)
* ext4: Fix corruption by hole punch in large files (Closes: #685726)
- rewrite punch hole to use ext4_ext_remove_space()
- fix hole punch failure when depth is greater than 0
- fix kernel BUG on large-scale rm -rf commands
* md: protect against crash upon fsync on ro array (Closes: #696650)
* net: Add alx driver for Atheros AR8161 and AR8162 (Closes: #699129)
- Mark as staging, since it has not been accepted upstream
* [armel/kirkwood] rtc-s35390a: add wakealarm support (Closes: #693997)
* [x86] i915: Invert backlight brightness control for various models
including Packard Bell NCL20 (Closes: #627372) and eMachines G725
(Closes: #680737)
- Also allow this behaviour to be enabled via module parameter
invert_brightness=1
* [amd64] edac: Enable EDAC_SBRIDGE as module (Closes: #699283)
* SCSI: Add virtio_scsi driver (Closes: #686636)
* [x86] sound: Enable LINE6_USB as module (Closes: #700211)
- Apply upstream changes up to Linux 3.8-rc1
* [armhf/mx5] Update description to mention i.MX53
* mm: Try harder to allocate vmemmap blocks (Closes: #699913)
* aufs: Update to aufs3.2-20130204:
- support for syncfs(2)
- possible bugfix, race in lookup
- bugfix, half refreshed iinfo
- possible bugfix, au_lkup_by_ino() returns ESTALE
* [x86] efi: Clear EFI_RUNTIME_SERVICES rather than EFI_BOOT by "noefi" boot
parameter
[ Aurelien Jarno ]
* [armhf/vexpress] Add kernel udebs.
[ Cyril Brulebois ]
* Bump python build-dep, needed since the switch from local SortedDict
to collections.OrderedDict (new in version 2.7).
[ Julien Cristau ]
* Backport drm and agp subsystems from Linux 3.4.29 (closes: #687442)
- [x86] i915: Fixes freezes on Ivy Bridge (Closes: #689268)
- nouveau: Support for newer nvidia chipsets (Closes: #690284)
- radeon: Support for HD7000 'Southern Islands' chips
- [x86] drm/i915: add Ivy Bridge GT2 Server entries (Closes: #684767)
- [x86] drm/i915: Close race between processing unpin task and queueing
the flip
--
Ben Hutchings
Sturgeon's Law: Ninety percent of everything is crap.
Attachment:
signature.asc
Description: This is a digitally signed message part