On Wed, 2004-04-14 at 09:04, Vincent Lefevre wrote: > BTW, I've just noticed another difference: I had kernel 2.4.25 at > this time (for the #1 -> #2 change). But as the default kernel is > 2.4.18 on my machine and I had to power the machine on, it booted > with this kernel 2.4.18, and the test with the second battery > occurred with this kernel.
Well, it would be useful if you could do a bit of analysis here, you have too much random facts for anything to be relevant. For example check that - 2.4.18: see if you can swap battery 1 in and out - same with 2.4.25 - 2.4.18: see if you can swap battery 2 in and out - same with 2.4.25 - do a few tests swapping from battery 1 to 2 and vice versa with both kernels Do enough tests to have some statistic, it may not fail all the time, could be only occasional. It's also possible that the problem is indeed the low-charge sleep kicking too late, but it could be the kernel not always properly shutting down some bit of HW Ben.
