Bug List for 2.2.0pre7 kernel

Alan Cox posts to linux-kernel:

Directly Exploitable

     * Procfs has locking errors on mm's in progress
     * Page table reclamation isnt done. Any user can toast the machine.

     * PPP Oopses with IPX
     * PPP breaks with Ricochet modems (-vj is a workaround)
     * PPP trace show driver breakages somewhere
     * Run two processes that keep rejoining multicast groups on an SMP
       box - (crash should be fixed by ANK by now)
     * Spam the box remotely with syn floods and other crap, it leaks
       memory (this takes a lot of crap to do fortunately)
     * select/poll magically break at some number of handles without an
       error (not a problem with default settings)
     * IDE probe often guesses wrong. Linux is impossible to install on
       these ranges of PCs. (most are now ok but not all -UMC problem)
     * isdn4linux is old not CVS version. If its not changing for 2.2.0
       include a note about the CVS one.
     * eata-dma driver crashes the machine if at any instant it cant grab
       atomic isa dma memory. (Possible fix mark it obsolete and use
       eata.c which works fine)
     * Video4linux bttv tends to crash machines grabbing 
        (update to follow tonight)
     * NTP: 2.2.x is missing the NTP updates still.
     * You can't mount an ext2fs cdrom. (Block size error). Works in 2.0
       Ok on IDE, breaks on SCSI
     * generic_file_mmap and MSDOS/UMSDOS disagree over who clears blocks
     * DaveM reports a pile of VMA operations done without locks held.
     * Need to review all CONFIG_EXPERIMENTAL tags
     * ide-scsi crashes (HP 8100 for one). Command translation problems
       can cause crashes.
     * SCSI error can hold the io lock, causing scsi interrupt handlers
       to need to sleep- then they panic since its an irq (1542 at least
     * Quota seems to crash some machines.
Unfixed but not vital

     * NFS client over tcp doesnt work (Now apparently fixed)
     * NFS readahead is too low
     * NFS performance to 8K page sized BSD boxes sucks rocks, 2.0.x is
       about 5 times faster for some cases
     * IPv6 calls set_multicast_list in the wrong context in progress
     * TCP fails to handle small SO_SNDBUF/RCVBUF settings
     * Make xconfig needs layout fixes patch exists
Fixed in -ac patches but not yet in Linus tree

For Linus:

     * AVL tree vm avoids bad perfomance problems
     * Certain numbers of scsi disks dont seem to work
     * Some further time fixes
     * JJ Swap header fix

     * Large file array support (will be required by vendors for several
       big name products). This is a tricky one. Im wearing too many hats
       to judge this objectively. Vendors will probably ship this anyway
       or something similar.
Linus doesnt want:

     * QlogicFC - no big problem, its seperate its clean and vendors can
       ship it and other driver addons easily as they do now. Its a
       nobrainer to install of the net.
     * Message queue support for 32bit sizes. This gets suprisingly
       tangled for compatibility. 2.3.x we have 64bit fileops and 32bit
       uid to face so do it all then.
     * IPI counter