Don't allow a backslash in a path component (CVE-2006-1863)
Bug #41284 reported by
Gary Coady
Affects | Status | Importance | Assigned to | Milestone | |
---|---|---|---|---|---|
linux-source-2.6.15 (Ubuntu) |
Fix Released
|
Medium
|
Unassigned |
Bug Description
[PATCH] Don't allow a backslash in a path component (CVE-2006-1863)
Unless Posix paths have been negotiated, the backslash, "\", is not a valid
character in a path component.
Signed-off-by: Dave Kleikamp <email address hidden>
Signed-off-by: Steve French <email address hidden>
Signed-off-by: Greg Kroah-Hartman <email address hidden>
To post a comment you must log in.
linux-source-2.6.15 (2.6.15-24.40) dapper-security; urgency=low
Changes by Ben Collins
* psmouse: Total reset for intellimouse. event_interrupt ible_timeout( ).
- Malone #30224
* powerpc: Add one-liner to fix physical memory mapping on some G3's.
- Malone #34508
* nsc-ircc: Update to latest code to fix crashes.
- Malone #46947
* nsc-ircc: Add some IBM thinkpads
* ahci: Add support for JMicron ahci controller.
- Malone #45839
* acx: Make all TI ACX111's use 1.2.1.34 firmware.
- Malone #30766
* sky2: Update to latest version, 1.4.
- Malone #38865 (and others)
* rt2500: Update to CVS code, which the maintainer says will fix our SMP
related bugs with this driver.
* PCI: reverse pci config space restore order. Stolen from upstream patch.
This should fix a few resume bugs for hardware that stricly needs to
adhere to PCI specs. MacTel is one of the biggest examples.
* Disable davicom usage in tulip driver to let dmfe module takeover.
- Malone #48287
* powerpc: Enable MESH and MAC_FLOPPY drivers.
* acpi/ec: Use semaphore instead of spinlock to get rid of missed interrupts
- Malone #39315
* i386/amd64: Change HZ=1000 to HZ=250. The high frequency was causing high
power consumption on some laptops, and also some latency under certain I/O
loads.
* irda/sir: Fix wait operations in kernel thread. Use proper
wait_
- Malone #45542
* hid-powerbook: Enable on i386 for MacTels.
* i386: Add Averatec 3200 to list of acpi=noirq dmi matches.
- Malone #48263
Changes by David S Miller
* [SPARC64]: Fix missing fold at end of checksums.
* [SPARC64]: Fix D-cache corruption in mremap.
* [TG3]: Handle Sun onboard tg3 chips more correctly.
Changes by Fabio M. Di Nitto
* [debian/config] Enable HUGE_TLB & Co. on sparc64.
Changes by Daniel T Chen
* sound/{ drivers/ opl3,synth/ emux}/: Fix port type bits {cs46xx, hda}/: Fix race in removing device
* sound/pci/ac97/: Add workaround for ASUS A6KM
* sound/pci/hda/: Fix handling of capture controls on ALC882 3/6-stack models
* sound/core/: Fix pcm-draining of capture stream in PCM middle layer
* sound/pci/hda/: Fix init verbs for ALC260 hp model
* sound/usb/: Add workaround for CSR Bluetooth Headphones
* sound/synth/emux/: Fix NULL pointer dereference
* sound/pci/hda/: Fix codec model for HP dc7600
* sound/pci/ice1712/: Don't use Consumer AC97 for Terratec DMX6fire
* sound/pci/hda/: Add support for more Sony Vaio models
* sound/pci/hda/: Add support for Sigmatel 922[7-9] HDA codecs
* sound/pci/ac97/: Add ThinkPad T41p to Jack Sense blacklist
* sound/pci/: Fix incorrect mixer element name for cmipci
* sound/pci/
* sound/pci/hda/: Add HP nx6320 to supported list
* sound/pci: Fix additional races in the irq handler and ioremap()
Changes by Ryan Lortie
* snd-hda-intel: fix routing on macbook
* usb-hid: enable Fn key on Macbook keyboard (Intel)
* libata: delay resume to wait for harddrives to spin up
* ich7-sci-en-quirk: poke the SCI_EN bit on Macbook resume
Security updates
* CVE-2006-1052: Cherry pic...