Very simply put, you first have to be able to access the Flash, e.g. run a Flash format test. If that works, you simply modify the exProductionBurn target...
11620
wangshh03
May 30, 2008 5:51 am
How can I download my application program to nor flash using pci bus without using jtag? Hardware reference information as follows: 1¡¢Main processor is...
11619
Alfred Lee
alfred17330
May 28, 2008 4:13 pm
Absolutely speaking, the lowest overhead is to allocate a large chunk of memory and manage yourself. You can take advantage of how you use your memory to save...
11618
Alfred Lee
alfred17330
May 28, 2008 4:07 pm
If I have to guess, it's the SMP that's the trouble. The ISR must be written much more carefully with proper spinlocks to work correctly on SMP machines....
11617
Chuck Peplinski
chuckpeplinski
May 28, 2008 2:09 pm
Of course, any memory allocation will have some overhead. malloc()/free() will have the lowest overhead. It has the corresponding drawback of lack of control...
11616
anton.denisov
May 28, 2008 9:14 am
... Îê. But if I need to allocate some part of memory without overhead what I have to do?...
11615
anton.denisov
May 28, 2008 9:14 am
... Îê. But if I need to allocate some part of memory without overhead what I have to do?...
11614
gu_jinbocn
May 28, 2008 5:32 am
I use sharedmemory to transfer video data from PC to PNX board. It run OK on Core2Duo platform running windows 2000 pro/server. But it will crash after 1 - 12...
11613
Harm Verhagen
harm_verhagen
May 27, 2008 9:59 pm
... The overhead is the biggest in the MpCached type memspace. (the shared memspace). The reason is that both the actual user buffer _and_ the administration...
11612
Chris Bore
chris_bore
May 27, 2008 2:24 pm
Following the poll on whether to participate in the Mississippi State University eGroup survey, we decided to do so. The survey, if you want to do it, is at: ...
11611
Alfred Lee
alfred17330
May 26, 2008 6:18 pm
I assume it doesn't cause any problem. You are simply getting more memory than you requested. So it is not a problem. I don't know the actual reason for...
11610
anton.denisov
May 26, 2008 12:46 pm
I found the answer: this overhead is used for containing a part of administration for the particular memory block. And size of this extra block of memory...
11609
anton.denisov
May 24, 2008 2:08 pm
I try to allocate some memory (size is about 4-10 b). But tmmlMalloc () returns pointer to 128 b ... Why? How can I repair it?...
11608
freeoicq
May 23, 2008 6:55 am
I have test dynboot on tm1500, I change _TMLINKTYPE to dynamic in tmdbg_jtag1500.bat, it seems ok, but when I download the image to the board, it crashed. ... ...
11607
gu_jinbocn
May 22, 2008 4:09 am
Test show: Windows XP is affected. windows 2003 may too. But Windows 2000 pro is OK, it seems windows 2000 shared memory no difference with normal memory. I'am...
11606
Alfred Lee
alfred17330
May 21, 2008 3:51 pm
This may be closer to what you intended? i = getc(image); while (i != EOF) tmDbg_Print (DBG_ROOT, DBG_ALL_FLAGS, "%c", i); i = getc(image); } Best Regards, ...
11605
Chuck Peplinski
chuckpeplinski
May 21, 2008 2:36 pm
um, your while loop is only over the print statement. You only read one character? If that's not it, are you running this on a PCI hosted system? Does ...
11602
hkr2000 hkr2000
freeoicq
May 16, 2008 1:18 am
Yeah, I think a cache copyback of all data that need to be copyback and a cache invalidate of all data that need to be invalidate should work, but I have not...
11601
Alfred Lee
alfred17330
May 15, 2008 4:21 pm
Ah, I see, you are the guy who ported Linux. So you must be talking about this: http://lxr.linux.no/linux/Documentation/memory-barriers.txt#L255 ...
11600
freeoicq
May 15, 2008 9:05 am
Perhaps I have not express what I mean. The problem is:when use rmb, all read request should be done, and when use wmb, all write request shoud be done, the...
11599
gu_jinbocn
May 15, 2008 12:30 am
PC CPU load increased. Sharedmemory return a Pointer which is considered as a normal memory pointer. but write to this region result longer time than a normal ...
11598
Alfred Lee
alfred17330
May 14, 2008 3:59 pm
May be we use different terms? There is cache copyback. After this instruction completes, dirty cache is forced written to memory. And then there is cache...
11597
Alfred Lee
alfred17330
May 14, 2008 3:55 pm
Is the CPU load increase on PC side or DSP side? Assuming DSP side, is it overall increase or increase in time spent reading the just written data? I believe...
11596
gu_jinbocn
May 14, 2008 9:39 am
I use shared memory to transfer video from Host PC to Trimedia board accross PCI bus. I find that: Data write to shared memory will significantly increase CPU...
11595
freeoicq
May 14, 2008 3:16 am
How to do a memory barrier, after which the data in cache and memory are identical?...
11594
freeoicq
May 6, 2008 5:58 am
I use usb over PCI in tm1500, and all hardware detect go corrent, but the pci int have some trouble. I use the INTA, and int the interrupt handler, it says: ...
11593
Harm Verhagen
harm_verhagen
May 2, 2008 8:25 pm
... Which is exactly what was happening. I found that tmdlMbs2 sometimes reports a done callback _before_ the corresponding task is actually finished. ...
11592
Harm Verhagen
harm_verhagen
May 2, 2008 2:38 pm
... function ... Thats exactly what I also thought. Its like those cache operations before/after painting the border are not happening. it seems those cache...
11591
Chuck Peplinski
chuckpeplinski
May 1, 2008 2:12 pm
Hi Harm. Logging this as a TCSHelp issue (rather than TMUG) will increase the chance that someone at NXP takes time to investigate. Logging here means that...
11590
Harm Verhagen
harm_verhagen
May 1, 2008 12:44 pm
Hi, I'm having some problems with a mixer sw component. Debugged for a while a started to doubt whether maybe hw might be the cause. I found one interesting...
