Another Pine64+ 2GB oddity with head -r323246 (so A64 in use): the occasional "Expensive timeout(9) function . . ."

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Another Pine64+ 2GB oddity with head -r323246 (so A64 in use): the occasional "Expensive timeout(9) function . . ."

Mark Millard-2
Another thing that is different from my prior -r308??? and
before Pine64+ use is that -r323246 gets occasional notices
like the following, even when sitting idle.

I've added a line identifying what the address just after "function: "
is (looked up via objdump on the kernel file).

Expensive timeout(9) function: 0xffff0000005dd044(0xffff00004091c000) 0.002561100 s
ffff0000005dd044 <awg_tick> str x21, [sp, #-48]!

Expensive timeout(9) function: 0xffff0000003978c8(0) 0.963402488 s
ffff0000003978c8 <pffasttimo> stp       x20, x19, [sp, #-32]!

Expensive timeout(9) function: 0xffff00000059f33c(0) 0.002390250 s
ffff00000059f33c <uma_timeout> stp      x22, x21, [sp, #-48]!

Expensive timeout(9) function: 0xffff00000021716c(0) 0.963501285 s
ffff00000021716c <newnfs_timer> adrp    x8, ffff000000983000 <wd_pretimeo_handle+0x30>

Expensive timeout(9) function: 0xffff0000003581b0(0) 0.963506661 s
ffff0000003581b0 <logtimeout> adrp      x8, ffff000000c85000 <dpcpu+0x158>

I do not see each of these every boot (and its later
operation). For example my most recent boot has only
gotten 0xffff00000021716c so far (somewhat after my
logging in as it turns out).

0xffff00000059f33c has happened during the "add net"
messaging during boot.



Some context notes follow.

This is with boot1.efs (as bootaa64.efi) based on having
done:

# svnlite update -r322941 /usr/src/sys/boot/efi/boot1/boot1.c /usr/src/sys/boot/efi/include/efiapi.h

so that what is built works on the Pine64+ 2GB.

This is with a debug kernel build (because a non-debug
kernel normally does not finish booting). (The debug
kernel sometimes fails to boot but usually does.)

This is with the kernel and world on mmcsd0 (because
USB is not working). No swap. (Historically I've had
world and swap on a USB SSD for the Pine64+ 2GB .)

This is with having dd'd:

# ls -lT /usr/local/share/u-boot/u-boot-pine64/*.bin
-rw-r--r--  1 root  wheel  505940 Sep  6 00:49:43 2017 /usr/local/share/u-boot/u-boot-pine64/u-boot-sunxi-with-spl.bin

that was built from:

# svnlite info /usr/ports/ | grep "Re[plv]"
Relative URL: ^/head
Repository Root: svn://svn.freebsd.org/ports
Repository UUID: 35697150-7ecd-e111-bb59-0022644237b5
Revision: 449313
Last Changed Rev: 449313

(The port had been updated after I'd last done such
a thing so I updated. It turns out that with
u-boot-sunxi-with-spl.bin no pine64_plus.dtb is
needed on mmcsd0s1 .)

===
Mark Millard
markmi at dsl-only.net

_______________________________________________
[hidden email] mailing list
https://lists.freebsd.org/mailman/listinfo/freebsd-arm
To unsubscribe, send any mail to "[hidden email]"