"Chris Antos" <[EMAIL PROTECTED]> wrote:
>from looking at the OS source for ResLoadForm (called by FrmInitForm), file
>"HwrRes.c" line 85, there's an obvious low-memory bug. it allocates memory
>but does not check for null. so if there's not enough memory, you'll get a
>priviledge violation trying to dereference null. check and see if the
>emulated device is nearly full (after 905k gremlins that's not hard to
>believe).
I tried to check memory, but I find that the IDE gets pretty unstable under
some low memory conditions. Running commands in the debugger console
yields unpredictable results.
I looked at the line you referred to in HwrRes.h and I see the call you're
referring to. Is this, in fact, a bug? Or should my programming never put
the OS in a condition where I expose this bug? I've learned to be
suspicious of my programming first and the OS second. ;-)
>this is the second pretty obvious low-memory bug that's been found in a
>couple weeks, so i'm sure the Palm guys are going to do a code review of all
>the OS code looking for low-memory bugs like this. or maybe they've already
>done so. unless they release a patch, it won't help anyone with existing
>devices, though.
Anyone from Palm care to comment on the release of a patch? I'd like to
know if I should stop sweating the consistent failure of Gremlins to reach
1,000,000 events, or if there are other things I should be doing to avoid
colliding with this OS issue.
And thanks for the help, Chris. Much appreciated. :-)
>> I'm in CurrentFORM and I want to get the text from a list
>(MyOtherFORMList)
>> in MyOtherFORM. MyOtherFORM isn't loaded so I've got to
>>
>> - load the form,
>> - get a pointer to the list,
>> - get a pointer to the listtext,
>> - copy the text,
>> - unload the form
[snip]
>> /**** PROBLEM HERE ****/
>> frm = FrmInitForm( MyOtherFORMForm );
[snip]
--
Andrew Ball
[EMAIL PROTECTED]
