Hello,
the command "soft_reset_halt" of the current git version of openocd segfault:
(gdb) bt full
#0 cortex_m3_assert_reset (target=0x818dba0) at cortex_m3.c:942
cortex_m3 = 0x818f118
swjdp = 0x818f1d0
reset_config =
jtag_reset_config =
retval =
On 17/12/2010, at 4:11 PM, Øyvind Harboe wrote:
> On Fri, Dec 17, 2010 at 7:06 AM, Peter Stuge wrote:
>> Øyvind Harboe wrote:
>>> Which is better? (whatever "better" is)
>>>
>>> This:
>>>
>>> if {[info exists CHIPNAME]} {
>>>set _CHIPNAME $CHIPNAME
>>> } else {
>>>set _CHIPNAME at91r4
Peter Stuge wrote:
> > Instead of:
> >
> > source [find target/foo.cfg]
> >
> > We could define a procedure:
> >
> > load_target foo at91r40008
>
> watch out for the general find target/* vs. a specialized
> load_target.
To be a little constructive here's an idea, but I dislike the name:
find
Øyvind Harboe wrote:
> > The shorter one. But it is rather confusing to deal with the
> > different types of Tcl variables (or whatever they are?) which
> > have the same names, but obviously do different things.
> >
> > E.g. what exactly is the difference between _CHIPNAME and $CHIPNAME ?
> >
> >
> The shorter one. But it is rather confusing to deal with the
> different types of Tcl variables (or whatever they are?) which
> have the same names, but obviously do different things.
>
> E.g. what exactly is the difference between _CHIPNAME and $CHIPNAME ?
>
> Would it be possible to get rid of
Øyvind Harboe wrote:
> >> set _CHIPNAME [defaultval $CHIPNAME at91r40008]
> >
> > Significantly better.
>
> It's trivial to do something shorter, like:
>
> defaultval CHIPNAME at91r40008
Much better still!
> The advantage of the more verbose syntax is that it
> is more self documenting:
>
> s
On Fri, Dec 17, 2010 at 7:06 AM, Peter Stuge wrote:
> Øyvind Harboe wrote:
>> Which is better? (whatever "better" is)
>>
>> This:
>>
>> if {[info exists CHIPNAME]} {
>> set _CHIPNAME $CHIPNAME
>> } else {
>> set _CHIPNAME at91r40008
>> }
>
> All these blocks in the target files have really
Øyvind Harboe wrote:
> Which is better? (whatever "better" is)
>
> This:
>
> if {[info exists CHIPNAME]} {
>set _CHIPNAME $CHIPNAME
> } else {
>set _CHIPNAME at91r40008
> }
All these blocks in the target files have really annoyed me.
> or this:
>
> set _CHIPNAME [defaultval $CHIPNAM
On 16.12.2010 16:58, Peter Stuge wrote:
> Michael Trensch wrote:
>> Probably it was an error to keep two branches and merge them by hand.
>>
>> I've made the changes in a temporary test environment and tried to merge
>> the stuff.
>> 1. cloned openocd's GIT repo
>> 2. git mv tcl/target/netx500.cfg
Michael Trensch wrote:
> I thought I had the Line-ending stuff in place:
> core.autocrlf=false
> core.safecrlf=true
Ok, should not do conversion then.
> Probably it was an error to keep two branches and merge them by hand.
>
> I've made the changes in a temporary test environment and tried to m
On 16.12.2010 16:39, Peter Stuge wrote:
> Hm. Did you actually rename the file, or copy/paste into a new file?
> Maybe the line endings got switched around; git usually has
> unix-style line endings for all files, and can do automatic
> conversion on checkout if desired. Once that starts getting us
Michael Trensch wrote:
> Good Hint, didn't know about that yet. As I told I am still a newbie
> to opensource projects.
No worries. You're doing fine.
> Mh I think I made an error in my local commits. When running -M -C the
> diff tells me i've renamed "netx500cfg --> hilscher_netx10.cfg
> (simi
On 16.12.2010 16:05, Peter Stuge wrote:
> Øyvind Harboe wrote:
>> Any objections?
> No. Keep in mind to create patches with -M -C when renaming or
> copying a file in a commit. Maybe some surgery would be good to make
> sure that the file rename is a rename in the repo. OTOH maybe it
> doesn't matt
On Thu, Dec 16, 2010 at 4:05 PM, Peter Stuge wrote:
> Øyvind Harboe wrote:
>> Any objections?
>
> No. Keep in mind to create patches with -M -C when renaming or
> copying a file in a commit. Maybe some surgery would be good to make
> sure that the file rename is a rename in the repo. OTOH maybe it
Øyvind Harboe wrote:
> Any objections?
No. Keep in mind to create patches with -M -C when renaming or
copying a file in a commit. Maybe some surgery would be good to make
sure that the file rename is a rename in the repo. OTOH maybe it
doesn't matter.
//Peter
Any objections?
--
Øyvind Harboe
Can Zylin Consulting help on your project?
US toll free 1-866-980-3434 / International +47 51 63 25 00
http://www.zylin.com/zy1000.html
ARM7 ARM9 ARM11 XScale Cortex
JTAG debugger and flash programmer
___
Openocd-d
Hi,
the attached patch adds additional support for the Hilscher netX chip
family.
But take care as it renames an existing script (target/netX500.cfg -->
target/hilscher_netX500.cfg).
Note that this is my first patch and if there's something wrong with it,
just tell me. I am still learning how to
Which is better? (whatever "better" is)
This:
if {[info exists CHIPNAME]} {
set _CHIPNAME $CHIPNAME
} else {
set _CHIPNAME at91r40008
}
or this:
set _CHIPNAME [defaultval $CHIPNAME at91r40008]
# Std proc to be added
proc defaultval {varname defval} {
if {[uplevel "info exists $var
18 matches
Mail list logo
