01:37
jclsn has quit [Ping timeout: 272 seconds]
01:37
vagrantc has quit [Quit: leaving]
01:39
jclsn has joined #u-boot
01:42
stgl has joined #u-boot
01:49
clamor has joined #u-boot
01:52
clamor has quit [Read error: Connection reset by peer]
02:36
mmu_man has quit [Ping timeout: 244 seconds]
03:04
vardhan has joined #u-boot
03:14
alpernebbi has quit [Ping timeout: 248 seconds]
03:15
persmule has quit [Remote host closed the connection]
03:29
alpernebbi has joined #u-boot
03:57
clamor has joined #u-boot
04:02
enok has joined #u-boot
04:08
KREYREN_ has quit [Remote host closed the connection]
04:09
KREYREN_ has joined #u-boot
04:45
clamor has quit [Ping timeout: 252 seconds]
04:46
clamor has joined #u-boot
04:58
goliath has joined #u-boot
05:29
enok has quit [Ping timeout: 265 seconds]
05:40
Forty-Bot has quit [Ping timeout: 272 seconds]
05:41
Forty-Bot has joined #u-boot
05:45
clarity has quit [Ping timeout: 248 seconds]
05:46
clarity has joined #u-boot
05:46
jfsimon1981_c has quit [Read error: Connection reset by peer]
05:46
jfsimon1981_c has joined #u-boot
05:50
monstr has joined #u-boot
05:55
monstr has quit [Ping timeout: 252 seconds]
06:09
ladis has joined #u-boot
06:13
pgreco_ has joined #u-boot
06:14
pgreco has quit [Ping timeout: 252 seconds]
06:44
jfsimon1981_c has quit [Ping timeout: 245 seconds]
06:45
umbramalison has joined #u-boot
06:45
umbramalison_alt has quit [Ping timeout: 272 seconds]
06:51
mckoan|away is now known as mckoan
06:51
<
mckoan >
good morning
07:02
enok has joined #u-boot
07:17
frieder has joined #u-boot
07:17
godvino has quit [Remote host closed the connection]
07:20
godvino has joined #u-boot
07:22
godvino has quit [Remote host closed the connection]
07:23
godvino has joined #u-boot
07:34
enok has quit [Ping timeout: 276 seconds]
07:35
frieder has quit [Ping timeout: 244 seconds]
07:48
frieder has joined #u-boot
07:49
ladis has quit [Ping timeout: 265 seconds]
07:50
ldevulder has joined #u-boot
07:50
ldevulder has quit [Remote host closed the connection]
08:00
ladis has joined #u-boot
08:01
enok has joined #u-boot
08:04
monstr has joined #u-boot
08:29
enok has quit [Ping timeout: 276 seconds]
08:37
naoki has quit [Quit: naoki]
08:37
naoki1 has joined #u-boot
08:39
naoki1 has quit [Client Quit]
08:39
naoki has joined #u-boot
08:47
matthewcroughan has joined #u-boot
08:48
matthewcroughan has quit [Client Quit]
08:48
mattcrgh has joined #u-boot
08:49
naoki has quit [Quit: naoki]
08:49
<
mattcrgh >
Does anyone know how I could reduce the time of u-boot from 2.5 seconds on rk3588? Most of the time seems to be spent on ATF in SPL
08:49
naoki has joined #u-boot
08:50
<
f_ >
mattcrgh: well, TF-A will need to stay for sure
08:50
<
mattcrgh >
Yeah annoyingly
08:50
<
mattcrgh >
I think the weird proprietary binaries are faster even because they fail at points
08:51
<
f_ >
in u-boot proper you can also reduce the pause time ("Press any key to stop autoboot")
08:51
<
f_ >
if you don't mind losing functionality Falcon mode might also be interesting to look at
08:51
<
mattcrgh >
Yeah I would like to, but not if I can only save 500ms
08:52
<
mattcrgh >
Proper only takes 500ms
08:52
<
f_ >
okay, then don't think there's that much you can do.
08:52
<
mattcrgh >
Yeah I just wanted to know if it was a hard lower limit due to the ATF
08:53
<
mattcrgh >
Do you think disabling memory would do much?
08:53
<
f_ >
SPL is the one that inits DRAM
08:54
<
mattcrgh >
Is there anything I can disable that could have the side effect of making ATF need to do less?
08:54
<
mattcrgh >
Even if it gimps the hardware, I don't need all these cores and ram
09:00
mattcrgh has quit [Quit: Client closed]
09:04
naoki has quit [Quit: naoki]
09:04
naoki has joined #u-boot
09:46
hsv has quit [Ping timeout: 252 seconds]
09:47
enok has joined #u-boot
09:48
naoki has quit [Ping timeout: 252 seconds]
09:48
hsv has joined #u-boot
09:52
mmu_man has joined #u-boot
10:00
mmu_man has quit [Ping timeout: 252 seconds]
10:01
mmu_man has joined #u-boot
10:06
rber|res has joined #u-boot
10:50
enok has quit [Ping timeout: 265 seconds]
11:09
enok has joined #u-boot
11:22
<
marex >
bring the hardware up in SPL ... takes some effort, but it is worth it
11:30
enok has quit [Ping timeout: 252 seconds]
11:34
clamor has quit [Read error: Connection reset by peer]
11:34
jfsimon1981_c has joined #u-boot
11:35
clamor has joined #u-boot
11:38
persmule has joined #u-boot
12:05
vardhan has quit [Remote host closed the connection]
12:05
vardhan has joined #u-boot
12:07
vardhan has quit [Remote host closed the connection]
12:09
trem has joined #u-boot
12:11
frieder has quit [Ping timeout: 252 seconds]
12:26
frieder has joined #u-boot
13:16
jybz has quit [Excess Flood]
13:16
jeeebz has joined #u-boot
13:29
sszy has joined #u-boot
13:29
enok has joined #u-boot
13:42
dhruvag2000 has joined #u-boot
14:00
goliath has quit [Quit: SIGSEGV]
14:01
haritz has quit [Remote host closed the connection]
14:16
mmu_man has quit [Ping timeout: 244 seconds]
14:23
mmu_man has joined #u-boot
14:34
mmu_man has quit [Ping timeout: 252 seconds]
14:38
jmasson has joined #u-boot
14:38
jmasson has left #u-boot [#u-boot]
14:57
mmu_man has joined #u-boot
15:19
dsimic has quit [Ping timeout: 252 seconds]
15:19
dsimic has joined #u-boot
15:21
enok has quit [Read error: Connection reset by peer]
15:23
mmu_man has quit [Ping timeout: 248 seconds]
15:24
dsimic has quit [Ping timeout: 268 seconds]
15:26
dsimic has joined #u-boot
15:35
jfsimon has quit [Remote host closed the connection]
15:39
jfsimon has joined #u-boot
15:43
goliath has joined #u-boot
15:52
dhruvag2000 has quit [Quit: Connection closed for inactivity]
15:56
frieder has quit [Remote host closed the connection]
16:03
rvalue has quit [Read error: Connection reset by peer]
16:04
rvalue has joined #u-boot
16:14
KREYREN_ has quit [Ping timeout: 264 seconds]
16:25
mmu_man has joined #u-boot
16:38
monstr has quit [Remote host closed the connection]
16:52
mckoan is now known as mckoan|away
17:16
Poltawer has joined #u-boot
17:16
enok has joined #u-boot
17:18
jfsimon1981_c has quit [Remote host closed the connection]
17:23
enok has quit [Ping timeout: 260 seconds]
17:35
vagrantc has joined #u-boot
18:45
enok has joined #u-boot
18:53
clamor has quit [Read error: Connection reset by peer]
18:53
clamor has joined #u-boot
19:04
clamor has quit [Read error: Connection reset by peer]
19:44
matthewcroughan has joined #u-boot
19:44
<
matthewcroughan >
marex I managed to use BOOTSTAGE=y to determine that `1,186,577 board_init_r`
19:44
<
matthewcroughan >
which means 1.1 seconds for board_init_r
19:45
<
matthewcroughan >
spl isn't where all the time is spent as it turns out
19:59
persmule has quit [Ping timeout: 264 seconds]
20:06
enok71 has joined #u-boot
20:07
enok has quit [Ping timeout: 252 seconds]
20:07
enok71 is now known as enok
20:12
enok has quit [Ping timeout: 252 seconds]
20:17
enok has joined #u-boot
21:16
trem has quit [Remote host closed the connection]
21:22
ladis has quit [Read error: Connection reset by peer]
22:06
enok has quit [Ping timeout: 248 seconds]
22:14
zear has quit [Ping timeout: 244 seconds]
22:19
vagrantc has quit [Ping timeout: 252 seconds]
22:33
goliath has quit [Quit: SIGSEGV]
22:42
<
marex >
matthewcroughan: huh
22:43
<
marex >
matthewcroughan: there is a way to insert markers at specific points, to narrow the delay down
22:43
<
marex >
BOOTSTAGE_MARKER() I guess ... try it in common/board_r.c
22:52
<
matthewcroughan >
I did end up doing this, and still wasn't able to figure it out, there's a mysterious 1.4s delay that isn't traceable via the INITCALL stuff
22:52
clever has joined #u-boot
22:53
<
matthewcroughan >
with clever :D
22:54
<
clever >
we had initially just shoved `bootstage_mark_name()` at random points in the init loops to bisect and try to narrow down the delay
22:56
<
clever >
marex: but it wasnt proving useful, so we then tried printing `timer_get_boot_us()` within the `INITCALL()`, and that revealed a 1.4 seconds delay before display_text_info (the first INITCALL?) and a 1 second delay between `run_main_loop` and `start_kernel`
23:05
zear has joined #u-boot
23:08
Poltawer has quit [Quit: WeeChat 4.6.2]
23:10
<
Tartarus >
I only noticed this because I was attempting to normalize how you start test.py and get it to generate and save the junitxml results from the tests.
23:25
Hypfer6 has quit [Ping timeout: 260 seconds]
23:37
<
matthewcroughan >
marex: if I use the proprietary vendor BL31, boot times are reduced by 400~ms but BOOTSTAGE is none the wiser
23:38
Hypfer6 has joined #u-boot
23:41
<
marex >
matthewcroughan: could it be that bootstage accounts for the SPL run time ?
23:41
<
marex >
matthewcroughan: i.e. vendor blob runs 400ms, SPL runs 1.2 seconds ?
23:41
<
marex >
if yes, dig into the SPL
23:42
<
marex >
or ... use DEBUG_UART and stick some printf("%d\n", _get_timer(0)); before the initcalls start , that will give you current timer start, which likely counts from 0 on boot
23:42
<
matthewcroughan >
marex: did exactly that but it doesn't start from 0
23:43
naoki has joined #u-boot
23:54
umbramalison_alt has joined #u-boot
23:55
umbramalison has quit [Ping timeout: 252 seconds]
23:59
jmasson has joined #u-boot