root/lm-sensors/branches/lm-sensors-3.0.0/TODO @ 4426

Revision 4426, 3.6 KB (checked in by khali, 8 years ago)

"make uninstall" is now implemented.
No driver for the MTP006.

  • Property svn:eol-style set to native
  • Property svn:keywords set to Author Date Id Revision
Line 
1LM-Sensors TODO list
2Contact us if you have comments or wish to help.
3------------------------------------------------
4
5KERNEL MODULES
6==============
7
8* Harmonize 'smart fan' support.
9  The following drivers support pwm. The ones that support
10  automatic or smart fan support are identified.
11  We have a /proc standard for manual PWM support that allows
12  for /proc based applications like pwmconfig and fancontrol
13  in our package.
14  Without a /proc standard for smart fan support such applications
15  are much more difficult.
16
17        Driver          Manual PWM      Smart fan
18        adm1026         standard        not supported
19        asb100          standard        ??
20        it87            non-std         yes
21        lm85            standard        yes
22        vt1211          doesn't work    not supported
23        w83627hf        standard        not supported
24        w83781d         standard        not supported (791 only)
25
26  The following drivers are for chips that have manual PWM only.
27
28        Driver          Manual PWM
29        mtp008          standard
30        smsc47m1        standard
31        vt8231          standard
32        w83781d (except 791d)   standard
33
34* Assistance on how to fix module refcounting is welcome.
35  To my understanding, it does not handle client->adapter->owner
36  correctly now, and it is possible to rmmod adapter while one of its
37  clients is in use.
38
39* ACPI and SMBus host
40 - ACPI subsystem may access SMBus host too. Locks?
41 - On my ancient board using i2c-via, suspend and power switch status
42   seems to be in the same register with SCL and SDA.
43
44* Suspend-to-disk, losing Vcc
45 - Reset i2c client to avoid alarms and SCI interrupts. Currently
46   sensors do not create interrupt with alarms.
47
48* Return from suspend
49 - Reinitialize i2c client.
50
51* Handle -1 returns from i2c layer rather than just delivering 0xFF to
52  /proc and libsensors; report errors through /proc 'alarms' or
53  new 'fail' entry? See adm1021 for a partial example.
54
55* ALL: cleanup_module is void; check also that cleaning up is done
56  properly, now we know that the module will be unloaded whatever we
57  do.
58
59* ALL chip drivers: add a readonly insmod option so we don't cause APM/ACPI
60  to go insane.
61
62* gl518sm: using iterate==2 causes a tread to be started. this thread remains
63  even after setting iterate==1 and/or removing the gl518sm module
64  <koenig@uranus.tat.physik.uni-tuebingen.de>
65
66* LM78 detection: Tom Webster has proven that the reset bit in the
67  id register can be one (strange!)
68
69* maxilife.c: Introduce new insmod variables
70
71* maxilife.c: Round REG_TO_VID correctly
72
73* Support 10-bit addresses. At this moment, they are supported nowhere, except
74  in Simon Vogl's i2c modules.
75
76* adm9240: check whether the current voltage computations are correct.
77  Probably not, as they are different from the datasheet specifications.
78  Also check for the supported dallas chip.
79
80* lm80: Check how OS alarms work. At page 17 of the sheet, it tells
81  something completely different from the description at page 25.
82
83
84LIBRARY
85=======
86
87* Easier use of /proc and libsensors without knowing the chip type
88  At the moment a programmer wishing to use the lm_sensors package needs to know
89  what chips are supported and needs to add all those chips into his package.
90  If you were to add a field or function to your structures to the effect of
91  sensor-function a programmer could use the package a lot more simply.
92  For values you could do something like -
93    enum sensor-type { volt, fan, temp, volt_min, volt_max, fan_div, fan_min,
94    temp_over, temp_hyst }
95
96* reload does not work (Bison/Flex problem?).
97
98* Some library routines are pretty inefficient right now.
99
100* Library should be split in more separate files, for better linking
101  behaviour.
102
103* Some adm9240 labels start with a digit; change this, it leads to confusing
104  syntax in the conf file (they have to be quoted)
Note: See TracBrowser for help on using the browser.