(BTW: I don't seem to have received the last two messages I sent to the list;
is there something wrong?)
> dmi: read: Success
This sucks. Dmidecode failed and I found out why (pretty lame bug indeed).
Patch attached.
Pam, do you have the time and knowledge to patch the dmidecode source that
comes with lm_sensors-2.6.5 (in prog/detect IIRC), reinstall it and rerun
sensors-detect? Could you also send us the output of dmesg and the output of
dmidecode itself, would be perfect :)
I also attached a simple perl script. Please run it as root (I swear it is
safe) and send me the output. It is an alternative method for IBM detection I
am working on. If it works, we can get rid of dmidecode really soon.
> Obviously I hit ctrl-c at that point.
There were no risk of breaking anything... Well, in theory ;)
Thanks A LOT for testing, be sure we appreciate :)
Khali.
-------------- next part --------------
--- dmidecode.c Fri Aug 9 23:26:31 2002
+++ dmidecode.c Sat Sep 28 10:52:43 2002
@@ -644,7 +644,14 @@
struct dmi_header *dm;
u8 *data;
int i=0;
+ int r=0, r2=0;
+ if(len==0)
+ {
+ fputs("dmi: no data\n", stderr);
+ return;
+ }
+
if(buf==NULL)
{
perror("dmi: malloc");
@@ -655,7 +662,9 @@
perror("dmi: lseek");
return;
}
- if(read(fd, buf, len)!=len)
+ while(r2!=len && (r=read(fd, buf+r2, len-r2))!=0)
+ r2+=r;
+ if(r==0)
{
perror("dmi: read");
return;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: ibm.pl
Type: application/x-perl
Size: 1647 bytes
Desc: not available
Url : http://lists.lm-sensors.org/pipermail/lm-sensors/attachments/20020928/87c060d1/attachment.bin
