memory as part of another structure. This rather large patch fixes that by
making the smaller piece an explicit member of the larger.
I brought this patch into sync with the latest SVN, and tested it by comparing
objects files from the tree pre- and post- patch.
The remaining 99% of the credit goes to Bob Schlärmann <bob2@dsv.nl> for
creating the original patch... thanks Bob!
* Bob Schlärmann <bob2@dsv.nl> [2006-12-24 18:29:39 +0100]:
> The following patch changes all of the lib/chips.c entries and any function
> that uses sensors_chip_feature. Most of the conversion was done through a
> perl script, so if you don't agree with the syntax it'll be easy to change.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@4287 7894878c-1315-0410-8ee3-d5d059ff63e0
* add -8
When using -Cf, flex defaults to generating a 7-bit scanner instead of 8-bit.
That's a bug, since we want the scanner to handle any character. Using -8 restores
the proper behavior.
* change -Cf to -Cfe
This allows a nice reduction in object size (from ~60KiB to ~15KiB) with almost
no impact to speed (adds < 100 cycles to scanner, where total cycles is > 1M).
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@4272 7894878c-1315-0410-8ee3-d5d059ff63e0
and 2) make the quoted escape sequence rules a little more compact. The first
cleanup is said to yield a slight speedup according to the flex manual - I
didn't bother to measure it.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@4236 7894878c-1315-0410-8ee3-d5d059ff63e0
uncovered three different bugs. This changeset fixes them all. It also
updates the sensors.conf(5) man page to be unambiguous regarding the
requirement for whitespace between elements.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@4235 7894878c-1315-0410-8ee3-d5d059ff63e0
First ran .../scripts/Lindent from Linux 2.6.x, then fixed up some
things that were particularly ugly. Verified the final result by
rebuilding and running 'objdump -d' on the before and after; there
were no differences.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@4073 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch makes optional the sysfs support in libsensors. The default is to
*disable* sysfs support if a 2.4.x kernel is detected, for all other versions,
support will be enabled. This can be overridden in the top-level Makefile.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3096 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch replaces the I2C chip enumeration (through sysfs) in lib/proc.c
with new functions in lib/sysfs (using libsysfs). The patch also updates
the init code accordingly.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3095 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch replaces the I2C bus enumeration (through sysfs) in lib/proc.c
with a new function in lib/sysfs (using libsysfs). The patch also updates
the init code accordingly.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3094 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch adds the files lib/sysfs.h and lib/sysfs.c, beginning with a rewrite
of the sysfs detection code (using libsyfs). In the process, I replaced two
poorly named globals with better names, and moved them to sysfs.c. The sysfs
detection now happens very early during library init, rather than later.
The long term plan is for the two global variables to disappear anyway, as
everything to do with sysfs should get moved into lib/sysfs.c.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3093 7894878c-1315-0410-8ee3-d5d059ff63e0
This trivial patch links the libsysfs library to the libsensors library.
One needs to build and install sysfsutils first:
http://linux-diag.sourceforge.net/Sysfsutils.html
This patch is naive, at least to start. Later it will be made optional
depending on whether support for Linux 2.6.x is needed, or just 2.4.x.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3092 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch moves two macros from lib/proc.c to lib/data.h. The macros
pick up the "sensors_" prefix along the way. The motivation is the
same as the previous patch.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3091 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch moves the chip and bus lists from lib/proc.c to lib/data.c.
The variables were already declared in lib/data.h. Motivation: when
support for kernel 2.4.x becomes optional, lib/proc.c may not even be
compiled into the library... but these variables are still needed.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@3090 7894878c-1315-0410-8ee3-d5d059ff63e0
(+5V) scaling
> I have been hunting down an issue related to the IN10 input of the
> LM93. I have observed that there is a rather large error between the
> actual voltage and the voltage level displayed by sensors(1) (~ .12V).
>
> Please find attached a small patch that fixes this problem. The data
> sheet is very ambiguous and I and another engineer made the same mistake
> calculating the voltage per DAC step the first time we read the spec sheet.
>
> The confusion stems from an entry in the "maximum voltage" column that
> is higher than the value in the "absolute maximum range" column. For 5V
> only, the value in the "register reading at maximum voltage" is for the
> voltage in the "absolute maximum range", NOT the "maximum voltage". Whew!
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2924 7894878c-1315-0410-8ee3-d5d059ff63e0
> [...] I've also included a simple patch to lm93.c to
> increase the wait between retries if there is a timeout.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2835 7894878c-1315-0410-8ee3-d5d059ff63e0
> Building (the real) 2.8.8 on RH7.3 uncovered a small C99-ism, patch
> against CVS is attached.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2746 7894878c-1315-0410-8ee3-d5d059ff63e0
* read dynamic VCCP error status for IN7 & IN8 (if we're in that mode)
* unmask #VRDHOT and dynamic VCCP (if we're in that mode) error events
* move 'start monitoring' command to the end of the inits
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2734 7894878c-1315-0410-8ee3-d5d059ff63e0
* VIDs were reading 10x too big - fixed integer magnitude
* enabling smart tach for one fan sometimes disabled it for others -
fixed bit-twiddling during writes
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2689 7894878c-1315-0410-8ee3-d5d059ff63e0
revision should be functionally equivalent with the previous, except one small
bugfix: a typo in one of the smart tach SYSCTL numbers. The unified diff from
this revision to the previous is nearly useless - sorry, it can't be helped.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2672 7894878c-1315-0410-8ee3-d5d059ff63e0
* reading #VRDHOT inputs
* reading GPIO pins (input only for now)
* reading alarms status
* #PROCHOT override control
* #PROCHOT sampling interval control
* Manual PWM output control
* PWM output frequency control
Also, some whitespace cleanup in the driver.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2660 7894878c-1315-0410-8ee3-d5d059ff63e0
standing false detection problem. Other changes in support of this custom
detection:
1) Changed pci_list to a hash, with keys of the form "1039:0016"
2) Added @kernel_version detection and comparison routine (thanks Khali)
3) Removed legacy sub: read_proc_pci
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2358 7894878c-1315-0410-8ee3-d5d059ff63e0
> Current gcc versions warn when type punning is used, as this breaks the
> strict aliasing analysis gcc 3.3 does by default. When checking the causes
> for these warnings I discovered that sensors_malloc_array,
> sensors_free_array, sensors_add_array_el and sensors_add_array_els in
> lib/general.c all take a 'void **' for the list. This is plain wrong as
> the C standard only allow 'char *' and 'void *' to alias any other type, so
> if these functions should take a generic pointer, it *must* be a 'void *'
> that's internally cast to a 'void **'.
>
> The attached patch does this and adapts all places where these functions are
> called. Note that when done correctly, no cast what-so-ever in the calling
> code is needed.
>
> The patch is against current lm_sensors2 cvs version.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2345 7894878c-1315-0410-8ee3-d5d059ff63e0
This patch fixes a "Trying to release non-existent resource" error that
occurs during rmmod when the device isn't actually present. It includes
some other cleanups too: error paths, whitespace, magic numbers, __devinit.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2248 7894878c-1315-0410-8ee3-d5d059ff63e0
The workarounds were obviated by Khali's patch which was merged
into Linus' tree at 2.6.1-bk6. Tested here on w83781d - would
appreciate testing on the other w83's...
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2237 7894878c-1315-0410-8ee3-d5d059ff63e0
stops short of full "support" for sensors, for now. I.e. all the
info is displayed, it's just not as pretty.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2206 7894878c-1315-0410-8ee3-d5d059ff63e0
chip, but we'll guess that GPIO5 are used for this purpose. It's
correct for (at least) Asus P4C800-E, and if it's wrong for any
others then we're no worse off. Also, switch to using common LM75
conversion functions.
git-svn-id: http://lm-sensors.org/svn/lm-sensors/trunk@2151 7894878c-1315-0410-8ee3-d5d059ff63e0