stolen has quit [Quit: Connection closed for inactivity]
qubuepe24 has quit [Remote host closed the connection]
qubuepe24 has joined #osdev
qubuepe24 has quit [Remote host closed the connection]
cloudowind has quit [Ping timeout: 252 seconds]
qubuepe24 has joined #osdev
qubuepe24 has quit [Remote host closed the connection]
<kazinsal>
geist: how's the storm treating you?
netbsduser` has joined #osdev
cloudowind has joined #osdev
netbsduser` has quit [Ping timeout: 252 seconds]
the_oz has quit [Read error: Connection reset by peer]
the_oz has joined #osdev
goliath has joined #osdev
<geist>
power has been out since about 6pm
<geist>
the PSE map is currently showing 350000 customers without power
<geist>
pretty epic
<cloudowind>
life stops withouth it hey
<geist>
this is pretty historic yeah
<geist>
showing over 550k in all of seattle area
craigo has joined #osdev
theruran has quit [Quit: Connection closed for inactivity]
GeDaMo has joined #osdev
spare has joined #osdev
Dead_Bush_Sanpa1 has joined #osdev
Dead_Bush_Sanpai has quit [Ping timeout: 252 seconds]
Dead_Bush_Sanpa1 is now known as Dead_Bush_Sanpai
annamalai has quit [Ping timeout: 276 seconds]
annamalai has joined #osdev
hwpplayer1 has joined #osdev
foudfou has quit [Remote host closed the connection]
<cloudowind>
thats millions of people without power , must be a serious issue
foudfou has joined #osdev
<klys>
power to the people, is the power corrupt
hwpplayer1 has quit [Remote host closed the connection]
<dinkelhacker>
geist: So the 150 ms is with caches turned off. By context switch I mean switching from one task to another. So basically both taks will just do a syscall in a loop, that will schedule the other task. The scheduler is a very simple round robin implementation. Also it might be noteworthy that the code runs from flash. As said once I turned on the i cache it was ~50 ms so about 3x faster. I assume
<dinkelhacker>
that if i increase the core's clock speed enabling the cache will have a greater effect.
<dinkelhacker>
I guess to really understand how fast it should be I need to find out how fast I should theoretically be able to get the instructions from the internal flash and calculate based on the 64MHz clock how much instructions I should be able to execute in that time. And if those 2000 task switches should be faster or not.
<dinkelhacker>
geist: So with caches on that means that 1 task switch takes about 25 us at 64MHz. That matches your expectation, right?
<dinkelhacker>
Out of curiosity I measured what's the cost of a single exception entry/exit with no scheduling. So immediately returning after the syscall... 13 us with caches off 6 us with i-cache enabled.
<the_oz>
trees to power lines: let me show you the dance of our peoples. *trust falls*
<the_oz>
power lines to trees: c'mre braNch let me show you the song of our peoples. ZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ oh no what'ya on fire fer?
<dinkelhacker>
the_oz: are you in the area of the storm?
<the_oz>
I am not, in pensacola fl. I have a friend that lives there so I heard about it
<dinkelhacker>
kk.. yeah seems to be pretty bad
LjL has quit [Read error: Connection reset by peer]