1 changes to Ken's file server to make this 63-bit file server
6 note: 2⁶=9,223,372,036,854,775,808 or 8EB (9.2×10ⁱ⁸)
8 • identified longs that refer to offsets, sizes and block numbers, and
9 changed them to type Off (vlong); fixed all print formats to match.
10 fixed byte-swapping for the 'x' config to match.
12 • fixed VLONG 9p1 message packing and unpacking macros to actually
13 handle 64-bit offsets and sizes.
15 • implemented triple-indirect blocks. affected code in
16 dev/cw.c port/con.c port/dentry.c port/sub.c
17 port/chk.c port/console.c port/portdat.h
19 • Fri Aug 6 16:50:59 PDT 2004
21 Plan 9 v4 63-bit file server
25 • added long(er) file name components (56 bytes), long enough for all but one
26 name in my /.longnames file (68-byte .xml name).
28 • Fri Aug 6 21:43:41 PDT 2004
30 Plan 9 v4 63-bit file server sizes
34 • touched up lib.h (from libc.h) to bring it up to date with formatting
35 functions, verbs & flags.
36 • check now reports stack usage: 320 bytes upon entry to fsck first time,
37 92 bytes of stack per recursion. given 16000 bytes of stack,
38 that's 170 recursions maximum.
39 • booted xtc (terminal) from fs64 (used fs64 as main file system)
41 note: current file server with triple-indirect blocks at 4k block size
42 has a maximum file size of ~505GB (5.42×10ⁱⁱ).
43 with quadruple-indirect blocks, max would be ~275TB @ 4k block size.
45 • got igbe fs driver working (a couple small changes)
46 • eliminated some gotos (started with 580, down to 454)
47 • added quadruple indirect blocks: lets us reach 2⁶ with a 32kB block size
48 • got igbe boot driver & pxe booting working
49 • on-disk qid paths are now Offs, but 9p1 qids on the wire are still ulongs
50 • generalised & parameterised indirect block implementation
51 • tested with plain w0 fs, cached fake worm on w0, cw jukebox (hp 160fx)
52 • ip directories in fs & fs64 are identical except for whitespace and
54 • replaced most of nemo's ide code with newer ide code lifted from 9load,
55 then from cpu kernel (sdata.c & support). this brings us dma, rwm & lba48,
56 finds ide controllers by itself, even pci ones, & copes with dead drives
57 (i.e., doesn't panic).
58 • fixed long-standing bug that caused a 5-second delay before each console
59 prompt on systems without a serial console.
60 • further type parameterisation: Userid (short), Timet (long), Devsize (vlong).
61 Comment on v7 kernel portability work, quoting scj & dmr from BSTJ v57
62 #6 part 2., p. 2038: ``The important data types used within the
63 system were identified and specified using typedef: disk offsets,
64 absolute times, internal device names, and the like. This effort was
65 carried out by K. Thompson.''
66 • corrected compat.h dependencies in mkfiles
67 • eliminated all warnings
68 • implemented truncation via wstat