1 |
wireless wrote: |
2 |
> Jason wrote: |
3 |
>> wireless wrote: |
4 |
>>> Mike Frysinger wrote: |
5 |
>>>> On Friday 12 September 2008, billium wrote: |
6 |
>>>>> get openocd from respository, svn checkout |
7 |
>>>>> svn://svn.berlios.de/openocd/trunk/openocd and extract |
8 |
>>>> why ? we have an openocd-9999.ebuild which fetches straight from the |
9 |
>>>> upstream site ... |
10 |
>>> I have to chuckle out loud...... |
11 |
>>> |
12 |
>>> |
13 |
>>> Some time ago, I inquired about using developing on embedded (gentoo) |
14 |
>>> linux using Eclipse and any sort of TAG/BDM device. |
15 |
>>> |
16 |
>>> |
17 |
>>> I still remember the (focused) scolding you gave me about how that |
18 |
>>> was not the *nix way. Command line, burn and churn, etc etc.... |
19 |
>>> |
20 |
>>> As an old (*nix) fart, I rather enjoyed (and agreed with) your prose. |
21 |
>>> However, the planet is being overrun with young kids that like their |
22 |
>>> gui-fied tools.... |
23 |
>>> |
24 |
>> |
25 |
>> Rhetorical Question: How many of those "young kids" are _good_? There |
26 |
>> is no shortcut to deep understanding (no, I'm no where close). |
27 |
>> Something a GUI obfuscates away, the cli celebrates. But then, we |
28 |
>> agree... ;-) |
29 |
> |
30 |
> Most of what we do is old fashion roll your own state machine, async |
31 |
> programming (8-32) bit processors for products. So any RTOS/executive is |
32 |
> a luxury.... The designs typically used >90% of the resources. |
33 |
> Tight and Spartan still rules the roost in our embedded world. |
34 |
> |
35 |
|
36 |
So, you're coming at it from the bottom up (really small -> linux) |
37 |
whereas I'm approaching it from the top down (desktop linux -> embedded |
38 |
linux). |
39 |
|
40 |
[snip] |
41 |
>>> Now you are pointing folks to openocd......? |
42 |
> |
43 |
>> Are you implying that openocd is a gui? I think you might be referring |
44 |
>> to something else. |
45 |
> |
46 |
> Eclipse vs command line.... look at the top of this page: |
47 |
> |
48 |
> http://www.yagarto.de/ |
49 |
> |
50 |
> Second bullet (*works with Elipse*). That's what I refer to a |
51 |
> gui-fied. |
52 |
|
53 |
[root@localhost] # for i in `equery files openocd`; do if [ -x ${i} -a ! |
54 |
-d ${i} ]; then echo "-->${i}"; ldd ${i}; echo |
55 |
"#############################################"; fi; done |
56 |
-->/usr/bin/openocd |
57 |
linux-gate.so.1 => (0xffffe000) |
58 |
libftd2xx.so.0 => /usr/lib32/libftd2xx.so.0 (0xf7f79000) |
59 |
libusb-0.1.so.4 => /lib32/libusb-0.1.so.4 (0xf7f70000) |
60 |
libdl.so.2 => /lib32/libdl.so.2 (0xf7f6c000) |
61 |
libc.so.6 => /lib32/libc.so.6 (0xf7e3e000) |
62 |
libpthread.so.0 => /lib32/libpthread.so.0 (0xf7e26000) |
63 |
/lib/ld-linux.so.2 (0xf7fb7000) |
64 |
############################################# |
65 |
|
66 |
_not_ gui. Just because one project (yagarto) uses openocd and works |
67 |
with eclipse, doesn't make openocd a gui. That kind of logic is just |
68 |
frightening. Are you sure you aren't a politician? |
69 |
|
70 |
[snip] |
71 |
>>> Maybe we can all agree on a BSP that works well with a JTAG and |
72 |
>>> Eclipse, and make it an option for Noobs wanting to get into |
73 |
>>> embedded Gentoo? I think this is an easier sell than GNAP....? |
74 |
>>> Keeping cost down is a good idea, and my early vote goes for |
75 |
>>> an Arm9 board. |
76 |
> |
77 |
>> I haven't needed a jtag for it, since the bootloader was sufficient, but |
78 |
>> I like the NSLU2 as a good starter kit. I recently saw one on ebay for |
79 |
>> ~$30US. I bought mine new years ago for ~$70US. |
80 |
> |
81 |
> OK, if the group decides not to want a JTAG, then that's cool with me. |
82 |
> Folks occasionally become interested in embedded gentoo, are open to |
83 |
> what processor/board to use. IMHO, it'd be very cool to have an |
84 |
> introduction to embedded gentoo, say like Das Blinkenlights, |
85 |
> or Das AD/DA, on a simple, low cost board. Ethernet is the only real |
86 |
> hardware requirement I'd suggest, so I'd be cool to whatever the |
87 |
> consensus is. |
88 |
|
89 |
I think your basic premise is off in that you're attempting to herd |
90 |
cats. It's faulty to assume there needs to be a consensus in the first |
91 |
place. Sounds very cathedral-ish. |
92 |
|
93 |
fwiw, to get where you want to go, I think the only thing g-e needs is a |
94 |
place for contributors to place howtos in a loosely standardized format |
95 |
(wiki? no, I didn't _say_ wiki ;-) ) A new user could then hit that |
96 |
site and try one that works for their end goal, eg firmware for you, |
97 |
low-power servers for me. |
98 |
|
99 |
> |
100 |
>> The next step after that (at least, what I did) was the Gateworks 2348-4 |
101 |
>> board. Same processor, 4 mini-pci, 2 ethernet, USB as an option. It's |
102 |
>> less than $300US. Openocd works with it's parallel jtag programmer. I |
103 |
>> haven't needed that yet either, since the bootloader works for me. |
104 |
> |
105 |
> Um, most of our customers want an custom bootloader. It's really not |
106 |
> an option when you write firmware for a custom designed board that |
107 |
> is going to become a product. Sure vendors provide reference code, but |
108 |
> I've never seen that be sufficient for a product. I know lots of |
109 |
> companies use code from the chip vendors (cisco), but, it gets them |
110 |
> into trouble, more often than most realize. Besides a JTAG can be used |
111 |
> for debugging lots of different firmware/hardware issues, particularly |
112 |
> when you have limited/constrained resources and many things flying |
113 |
> around, asynchronously. You could not even get a firmware engineer |
114 |
> anywhere I've seen that would do a project, without a JTAG, BDM or |
115 |
> ICE. This is where embedded linux (and hopefully embedded gentoo) can go |
116 |
> to whole new level, when you customize device drivers for various |
117 |
> optimized needs, under constrained resources. Am I talking about |
118 |
> deviation from what the kernel gods have published? Sure, but only |
119 |
> in small amounts, and it is done quite often with embedded products. |
120 |
> Just look at the number of vendors now offering 'embedded linux'. |
121 |
> |
122 |
|
123 |
This is a different flavor of the top-down/bottom-up discussion. We |
124 |
have different needs. As long as the bootloader launches the linux |
125 |
kernel with the correct board id and command line arguments, I'm happy. |
126 |
Caveat that with no crypto/signature/lockdown crap... |
127 |
|
128 |
> |
129 |
>> The ethernet support is native in the vanilla kernel since about 2.6.21 |
130 |
>> or .23. As of 2.6.27, the kernel has support for the hardware crypto |
131 |
>> engine. IPsec VPN gateway, anyone? |
132 |
> |
133 |
> I'm not trying to upset anyone, I just see things slightly different |
134 |
> than most, because, as the end of the battle, I do want linux as a |
135 |
> workstation development platform (complete with a killer IDE) and |
136 |
> embedded linux (customized) as a ~rtos to win, or at least be a |
137 |
> realistic option for many processors and new products. |
138 |
> Sure (linux) is gaining market share, but, I think that |
139 |
> is a (relatively) few products that hit very high volume. It certainly |
140 |
> does not dominate where most firmware engineers actual do their work, |
141 |
> imho. |
142 |
|
143 |
I'm confused. Are you looking to use linux (g-e) to develop bare metal |
144 |
firmware, or to develop embedded linux firmware? |
145 |
|
146 |
Jason. |