1 ## Note: this is my personal todo and ideas list (alexh@)
3 * add a communication channel mechanism to dm
4 - essentially a way to send messages and receive responses to dm target instances
7 - Improve to support external scripts/programs providing passphrases
10 - Add plain64 iv support
11 - Think about benbi iv support
12 - Think about cascaded ciphers (a la TrueCrypt)
14 * Add TrueCrypt support
15 - Simply add a userland tool that sets dm_target_crypt up with the right parameters, extracted from the TrueCrypt header
19 * Port hwpmc & dig into (boot-up) performance
23 - separate out common arch parts (linprocfs, for example)
25 * Fix the crash analysis script (or rather the programs it calls [some segfault])
29 * Take a look at updating lvm/dm/libdm
32 o Added VT6105M specific register definitions. VT6105M has the
33 following hardware capabilities.
34 - Tx/Rx IP/TCP/UDP checksum offload.
35 - VLAN hardware tag insertion/extraction. Due to lack of information
36 for getting extracted VLAN tag in Rx path, VLAN hardware support
37 was not implemented yet.
38 - CAM(Content Addressable Memory) based 32 entry perfect multicast/
41 o Implemented CAM based 32 entry perfect multicast filtering for
42 VT6105M. If number of multicast entry is greater than 32, vr(4)
43 uses traditional hash based filtering.
45 * rip out the disk partitioning from the disk subsystem and implement it in a more general fashion
46 - crazy idea: as dm targets with an auto-configuration option!
48 * sync some more opencrypto from OpenBSD
50 * ATA (automatic) spindown (see FreeBSD current)
53 http://svn.freebsd.org/viewvc/base?view=revision&revision=127969
57 - some incorrect accounting going on, don't remember details :)
60 - make it work without whiteout
67 * RedZone, a buffer corruption protection for the kernel malloc(9) facility has been implemented.
68 - This detects both buffer underflows and overflows at runtime on free(9) and realloc(9),
69 and prints backtraces from where memory was allocated and from where it was freed.
72 * port uart driver (?)
74 * port wscons (?) or update syscons
75 - probably way too much effort (wscons)
79 - wrapper is included for userland; should be easy to port
80 - http://svn.freebsd.org/viewvc/base?view=revision&revision=184610
81 - http://turbocat.net/~hselasky/usb4bsd/
82 - http://gitweb.dragonflybsd.org/~polachok/dragonfly.git/shortlog/refs/heads/usb2
84 * suspend/resume for SMP x86
85 - http://lists.freebsd.org/pipermail/freebsd-acpi/2008-May/004879.html
87 * AMD64 suspend/resume
88 - http://svn.freebsd.org/viewvc/base?view=revision&revision=189903
95 [alexh@leaf:~/home] $ roundup-server -p 8080 bt=bugtracker
98 -05:48- : dillon@: no, double frees to the object cache are nasty. It can't detect them. the object
99 winds up in the magazine array twice
100 -05:48- : dillon@: (and possibly different magazines, too)
101 -05:49- : alexh@: can't I just write some magic to a free object on the first objcache_put and check
102 if it's there on objcache_put?
103 -05:49- : alexh@: and clear it on objcache_get, anyways
104 -05:50- : dillon@: no, because the object is still may have live-initialized fields
105 -05:50- : dillon@: because it hasn't been dtor'ed yet (one of the features of the objcache, to avoid
106 having to reinitialize objects every time)
107 -05:50- : dillon@: the mbuf code uses that feature I think, probably other bits too
108 -05:51- : dillon@: theoretically we could allocate slightly larger objects and store a magic number at
109 offset [-1] or something like that, but it gets a little iffy doing that
110 -05:52- : dillon@: the objcache with the objcache malloc default could probably do something like that
112 -05:52- : dillon@: I don't consider memory tracking to be a huge issue w/ dragonfly, though I like the
113 idea of being able to do it. It is a much bigger problem in FreeBSD due to the
114 large number of committers
117 -05:55- : dillon@: For the slab allocator you may be able to do something using the Zone header.
118 -05:55- : dillon@: the slab allocator in fact I think already has optional code to allocate a tracking
119 bitmap to detect double-frees
120 -05:56- : dillon@: sorry, I just remembered the bit about the power-of-2 allocations
121 -05:56- : dillon@: for example, power-of-2-sized allocations are guaranteed not only to be aligned on
122 that particular size boundary, but also to not cross a PAGE_BOUNDARY (unless the
124 -05:57- : dillon@: various subsystems such as AHCI depend on that behavior to allocate system
125 structures for which the chipsets only allow one DMA descriptor.
126 -05:59- : alexh@: http://svn.freebsd.org/viewvc/base/head/sys/vm/redzone.c?view=markup&pathrev=155086
127 < this is redzone. it basically calls redzone_addr_ntor() to increase the size in
128 malloc(), and then redzone_setup() just before returning the chunk
129 -06:02- : dillon@: jeeze. that looks horrible.
130 -06:03- : alexh@: I don't quite get that nsize + redzone_roundup(nsize)
131 -06:03- : dillon@: I don't get it either. It would completely break power-of-2-sized alignments in the
133 -06:04- : dillon@: hmmm. well, no it won't break them, but the results are oging to be weird
134 -06:04- : dillon@: ick.
136 -06:15- : dillon@: if the original request is a power of 2 the redzone adjusted request must be a power
138 -06:15- : dillon@: basically
139 -06:16- : dillon@: so original request 64, redzone request must be 128, 256, 512, 1024, etc.
140 -06:16- : alexh@: yah, k
141 -06:16- : dillon@: original request 32, current redzone code would be 32+128 which is WRONG.
142 -06:16- : alexh@: how big is PAGE_SIZE ?
143 -06:16- : dillon@: 4096 on i386 and amd64
144 -06:17- : alexh@: and one single malloc can't be bigger than that?
145 -06:17- : dillon@: I'm fairly sure our kmalloc does not guarantee alignment past PAGE_SIZE (that is,
146 the alignment will be only PAGE_SIZE eve if you allocate PAGE_SIZE*2)
147 -06:17- : dillon@: a single kmalloc can be larger then PAGE_SIZe
148 -06:18- : dillon@: it will use the zone up to around 1/2 the zone size (~64KB I think), after which it
149 allocates pages directly with the kernel kvm allocator
150 -06:18- : dillon@: if you look at the kmalloc code you will see the check for oversized allocations
151 -06:18- : alexh@: yah, saw that
152 -06:18- : alexh@: "handle large allocations directly"
153 -06:19- : alexh@: not sure how to do this, really, as the size is obviously also changed in
155 -06:20- : alexh@: but kmem_slab_alloc isn't called always, is it?
156 -06:20- : alexh@: only if the req doesn't fit into an existant zone
157 -06:20- : dillon@: right
158 -06:20- : dillon@: you don't want to redzone the zone allocation itself