Summary of changes from v2.4.33.3 to v2.4.33.4 ============================================ Akinobu Mita (1): WATCHDOG: sc1200wdt.c pnp unregister fix. dann frazier (1): Backport fix for CVE-2006-4997 to 2.4 tree Dick Streefland (1): incorrect timeout in mtd AMD driver of 2.4 kernel Geert Uytterhoeven (1): fbdev: correct buffer size limit in fbmem_read_proc() Herbert Xu (1): SCTP: Always linearise packet on input Jeff Garzik (1): ISDN: fix drivers, by handling errors thrown by ->readstat() Jurzitza, Dieter (1): really fix size display for sun partitions larger than 1TByte Martin Schwidefsky (3): copy_from_user information leak on s390. s390 : fix typo in recent copy_from_user fix [S390] user readable uninitialised kernel memory (3rd version) Michael Chen (1): i386: fix overflow in vmap on an x86 system which has more than 4GB memory. mostrows@earthlink.net (1): Advertise PPPoE MTU NeilBrown (1): knfsd: Fix race that can disable NFS server. Patrick McHardy (1): [NETFILTER]: Fix deadlock on NAT helper unload PaX Team (2): MIPS: fix long long cast in pte macro i386: fix long long cast in pte macro Pete Zaitcev (2): USB: Little Rework for usbserial USB: unsigned long flags Steffen Maier (1): block: fix negative bias of ios_in_flight (CONFIG_BLK_STATS) because of unbalanced I/O accounting Stephen Hemminger (1): [BRIDGE]: netfilter deadlock Toyo Abe (1): x86_64: Fix missing delay when the TSC counter just overflowed Willy Tarreau (10): i386: remove unsigned long long cast in __pte() macro. 2.4.x: i386/x86_64 bitops clobberings [PATCH-2.4] i2c-elv: fix erroneous '&&' operator fix "&& 0xffff" typo in gdth.c fix obvious "&& 0xFFFFFF" typo in cpqfcTSworker fix "&& 0xff" typo in qeth_qdio_input_handler fix two "&& 0x03" in usbnet EXT3: avoid crashing by not dividing by zero. EXT2: avoid crashing by not dividing by zero. Change VERSION to 2.4.33.4