This problem never existed in Vista in the first place. It’s been fixed in Linux since 177.73 and Windows since 178.08. If you’re encountering some other failure, send me a small repro case.
Hate to revive a long-dead thread, but this looks to be the exact same error I’m encountering… which seems odd considering a new driver has come out since then and the error theoretically never existed on Vista to begin with? Basically what I’m doing is this:
[codebox]while(blah)
{
for(#rows being processed)
allocate the size of each row;
do some stuff with all the rows;
for(#rows being processed)
free those rows;
}[/codebox]
With my current test data, the program consistently fails on the first row of the 2nd time through the while. For this program, it is not possibly to just initialize all the rows to the max possible length, as that could theoretically be any arbitrarily, very high size… but in reality it’s usually something like ~10. The specific call that’s failing in this test case is only trying to allocate 520 bytes.
Is this possibly still the result of the aforementioned “known (but fixed) bug”, or does someone have an idea for another cause?
Many thanks in advance.