Linux 2.4.21-rc1

[Marcelo Tosatti]

Here goes the first candidate for 2.4.21.

Please test it extensively.

Summary of changes from v2.4.21-pre7 to v2.4.21-rc1
============================================

<[e-mail]>:
  o USB: patch for oops in io_edgeport.c

<[e-mail]>:
  o USB: Patch against unusual_devs.h to enable Pontis SP600

<[e-mail]>:
  o USB: uhci bandaid

<[e-mail]>:
  o USB: ftdi_sio update

<[e-mail]>:
  o [SCTP/IPV6]: Move sockaddr storage and in6addr_{any,loopback} to generic places

<[e-mail]>:
  o [ATM]: Make ia64 include ATM driver config

<[e-mail]>:
  o [ATM]: Get lec net_device names correct
  o [ATM]: Obsolete some atm_vcc members
  o [ATM]: Fix idt77252/sch_atm/pppoatm compilation
  o [ATM]: cleanup nicstat, suni and idt77105
  o [ATM] nicstar doesnt count all dropped pdus and powerpc fixup
  o [ATM] s/uni driver overwrites 8-/16-bit mode
  o [ATM]: Fix total_len calculation in IPHASE driver
  o [ATM]: Fix IPHASE build with debugging enabled

<[e-mail]>:
  o [IPV6]: Add anycast support

<[e-mail]>:
  o [NETFILTER]: Fix modify-after-free bug in ip_conntrack

<[e-mail]>:
  o [NETFILTER]: Fix ipfwadm_core.c compile failure
  o [NETFILTER IPV6]: Fix Makefile typo

<[e-mail]>:
  o [VLAN]: Fix memory leak in procfs handling

<[e-mail]>:
  o USB: scanner.c endpoint detection fix

<[e-mail]>:
  o [NETFILTER]: iptables iptable_mangle LOCAL_IN bugfix
  o [NETFILTER]: ipt_REJECT bugfix for TCP RST packets + asymm. routing

<[e-mail]>:
  o USB: New USB serial device ID: Asus A600 PDA cradle

<[e-mail]>:
  o [NETFILTER]: Add tftp conntrack + NAT support

<[e-mail]>:
  o [IPV6]: Allow protocol to percolate up into rt6 routing operations

<[e-mail]>:
  o [NETFILTER]: Add amanda conntrack + NAT support

<[e-mail]>:
  o [TCP]: Missing SNMP stats

<[e-mail]>:
  o [NETFILTER]: ip_conntrack bugfix for LOCAL_NAT and PPTP

<[e-mail]>:
  o Fix kunmap_atomic debugging problem

<[e-mail]>:
  o [ATM]: Compile fix for net/atm/br2684.c

<[e-mail]>:
  o USB: enable Motorola cellphone USB modems

<[e-mail]>:
  o Fix ac97 incomplete headers

<[e-mail]>:
  o [IPV6]: Use RFC2553 constant variable

Adrian Bunk <[e-mail]>:
  o [NF/IPV6]: Remove all ipv6_ext_hdrs from ip6tables
  o [ATM]: Fix IPHASE driver build
  o Fix aic7xxx compilation

Alan Stern <[e-mail]>:
  o USB: usb-storage START-STOP under Linux 2.4

Alexey Kuznetsov <[e-mail]>:
  o [IPV4]: Fix deadlock in IGMP locking
  o [IPV6]: Correct CHECKSUM_HW handling in tcp_v6_send_check

Andi Kleen <[e-mail]>:
  o x86-64 update

Andreas Dilger <[e-mail]>:
  o don't allocate/free blocks in system areas

Andries E. Brouwer <[e-mail]>:
  o compilation fix for 2.4.21-pre7
  o Fix SCSI size reporting

Ben Collins <[e-mail]>:
  o IEEE-1394/Firewire update

Benjamin Herrenschmidt <[e-mail]>:
  o PPC32: Do better cache flushes around L2 cache ctrl register changes
  o PPC32: Factor out common code for saving/restore CPU special-purpose registers, used on SMP and for sleep/wakeup.
  o PPC32: Make sure IPI handlers run with interrupts disabled
  o PPC32: Add proper /proc/ide entry for pmac
  o PPC32: Update ide-pmac driver

Christoph Hellwig <[e-mail]>:
  o [NETFILTER]: 2.4 firewalling compat code removal
  o [NET]: Backport generic fc_type_trans to 2.4

David Brownell <[e-mail]>:
  o USB: ehci-hcd, minor hardware tweaks
  o USB: usbcore deadlock paranoia
  o USB: CDC Ether fix notifications

David S. Miller <[e-mail]>:
  o [IPV6]: Undo __constant_{n,h}to{n,h}l from anycast patch
  o [SPARC64]: Fix trap stack allocations so gcc-3.x builds work
  o [SCHED]: Some schedulers forget to flush filter list at destroy
  o [PKTSCHED]: Fix double-define of __inline__ et al
  o [IP TUNNEL]: inet_ecn_decapsulate modifies bits in wrong header
  o [PKT_SCHED]: Remove ugly arch ifdefs from generic code
  o [NETFILTER IPV6]: Fix route leak in ip6_route_me_harder

Geert Uytterhoeven <[e-mail]>:
  o Amiflop mod_timer()
  o Duplicate PROC_CONSOLE()
  o 2.4 IDE core code for m68k
  o 2.4 IDE driver code for m68k
  o M68k raw I/O updates
  o Generic RTC driver
  o M68k ndelay()
  o M68k needs WANT_PAGE_VIRTUAL

Hideaki Yoshifuji <[e-mail]>:
  o [IPV6]: Use "const" qualifier
  o [IPV6]: Use ipv6_addr_any() for testing unspecified address
  o [IPV6]: Don't allow multiple instances of the same IPv6 address on an interface
  o [IPV6]: Set noblock to 1 in NDISC sock_alloc_send_skb calls

James Morris <[e-mail]>:
  o [NET]: dst_clone --> dst_hold where appropriate
  o [PKTSCHED]: Kill redefinition of IPPROTO_ESP in sch_sfq.c

Jens Axboe <[e-mail]>:
  o Fix ide request races which resulted in corruption

Marcelo Tosatti <[e-mail]>:
  o Cset exclude: [e-mail]|ChangeSet|20030417235935|56567
  o Add missing HPT366 ID
  o Updated EXTRAVERSION to -rc1

Mark A. Greer <[e-mail]>:
  o PPC32: Add support for SERIAL_IO_PORT ports to the gen550 backend

Mikael Pettersson <[e-mail]>:
  o fix dmi_scan breakage
  o fix APIC bus errors on SMP K7 boxes in UP mode

Oleg Drokin <[e-mail]>:
  o reiserfs: Fix recenly introduced journal sanity check that breaks replay on old filesystems
  o reiserfs: Fix for journal replay process, to only replay transactions from last mount. By Chris Mason

Oliver Neukum <[e-mail]>:
  o Honour HFS lock bits

Paul Mackerras <[e-mail]>:
  o PPC32: Fix the interrupt entry path for POWER3 processors
  o PPC32: Clean up arch/ppc/mm/Makefile a little
  o PPC32: xmon fixes for CHRP, powerbooks, and SMP systems
  o PPC32: fix indentation in include/asm-ppc/bootinfo.h
  o PPC32: Restructure the top-level interrupt handling loop
  o PPC32: Align boot wrapper data segment on page boundary
  o PPC32: Make readb/w/l completely synchronous

Petko Manolov <[e-mail]>:
  o USB: pegasus link status detection fix

Randy Dunlap <[e-mail]>:
  o [NET]: typo and comment fixes

Randy Dunlap <[e-mail]>:
  o update unexpected IO APIC detection

Rusty Russell <[e-mail]>:
  o Fix minor NAT parsing issue

Stephen C. Tweedie <[e-mail]>:
  o 2.4: Fix for jbd compiler warnings

Tom Rini <[e-mail]>:
  o PPC32: Actually fix KGDB like Mark Greer mentioned
  o PPC32: Remove an option to partically disable the d-cache