Re: kernel: warning: total configured swap (493567 pages) exceeds maximum recommended amount

On Nov 11, 2025, at 10:53, bob prohaska <[email protected]> wrote:

> On Tue, Nov 11, 2025 at 05:09:46PM +0000, void wrote:
>> 
>> I've been testing with
>> 
>> % fio --name=random_rw_test --filename=./testfile1 --rw=randrw --bs=128k \
>> --ioengine=posixaio --iodepth=256 --numjobs=4 --runtime=120 --time_based \
>> --group_reporting --direct=1 --size=1G
>> 
>> I'd be interested to know what the last few lines after "Run status group 0
>> (all jobs):" is for your usb2 spinning context
>> as I have no usb2 interfaced hard drive i can test with
>> -- 
>> 
> 
> My Pi2's are busy for now, but an idle Pi4 (so, USB3) reported
> 
> bob@nemesis:~ % fio --name=random_rw_test --filename=./testfile1 --rw=randrw --bs=128k \
> ? --ioengine=posixaio --iodepth=256 --numjobs=4 --runtime=120 --time_based \
> ? --group_reporting --direct=1 --size=1G
> random_rw_test: (g=0): rw=randrw, bs=(R) 128KiB-128KiB, (W) 128KiB-128KiB, (T) 128KiB-128KiB, ioengine=posixaio, iodepth=256
> ...
> fio-3.34
> Starting 4 processes
> random_rw_test: Laying out IO file (1 file / 1024MiB)
> Jobs: 1 (f=1): [_(1),m(1),_(2)][66.1%][r=9306KiB/s,w=10.5MiB/s][r=72,w=83 IOPS][eta 01m:05s]     
> random_rw_test: (groupid=0, jobs=4): err= 0: pid=12593: Tue Nov 11 10:45:05 2025
>  read: IOPS=83, BW=10.4MiB/s (10.9MB/s)(1320MiB/126934msec)
>    slat (usec): min=2, max=326, avg= 4.48, stdev= 5.74
>    clat (msec): min=361, max=32683, avg=6057.95, stdev=2470.88
>     lat (msec): min=361, max=32683, avg=6057.95, stdev=2470.88
>    clat percentiles (msec):
>     |  1.00th=[  372],  5.00th=[ 1301], 10.00th=[ 2534], 20.00th=[ 4044],
>     | 30.00th=[ 5134], 40.00th=[ 5873], 50.00th=[ 6611], 60.00th=[ 7013],
>     | 70.00th=[ 7483], 80.00th=[ 7819], 90.00th=[ 8221], 95.00th=[ 8658],
>     | 99.00th=[13624], 99.50th=[14295], 99.90th=[15368], 99.95th=[15503],
>     | 99.99th=[17113]
>   bw (  KiB/s): min= 2548, max=124672, per=100.00%, avg=43049.66, stdev=9073.48, samples=243
>   iops        : min=   17, max=  972, avg=334.08, stdev=71.00, samples=243
>  write: IOPS=86, BW=10.8MiB/s (11.3MB/s)(1374MiB/126934msec); 0 zone resets
>    slat (usec): min=9, max=1141, avg=35.08, stdev=20.94
>    clat (msec): min=361, max=19451, avg=5676.77, stdev=2092.45
>     lat (msec): min=361, max=19451, avg=5676.80, stdev=2092.45
>    clat percentiles (msec):
>     |  1.00th=[  368],  5.00th=[ 1301], 10.00th=[ 2534], 20.00th=[ 3842],
>     | 30.00th=[ 4732], 40.00th=[ 5336], 50.00th=[ 6409], 60.00th=[ 6745],
>     | 70.00th=[ 7148], 80.00th=[ 7550], 90.00th=[ 7886], 95.00th=[ 8020],
>     | 99.00th=[ 8658], 99.50th=[ 8792], 99.90th=[10000], 99.95th=[11073],
>     | 99.99th=[12953]
>   bw (  KiB/s): min= 1517, max=140152, per=100.00%, avg=54330.03, stdev=11678.97, samples=201
>   iops        : min=    8, max= 1092, avg=422.08, stdev=91.30, samples=201
>  lat (msec)   : 500=1.13%, 750=1.15%, 1000=1.19%, 2000=4.67%, >=2000=91.86%
>  cpu          : usr=0.10%, sys=0.49%, ctx=19290, majf=0, minf=4
>  IO depths    : 1=0.1%, 2=0.1%, 4=0.1%, 8=0.1%, 16=0.3%, 32=2.0%, >=64=97.4%
>     submit    : 0=0.0%, 4=100.0%, 8=0.0%, 16=0.0%, 32=0.0%, 64=0.0%, >=64=0.0%
>     complete  : 0=0.0%, 4=97.7%, 8=0.8%, 16=0.5%, 32=0.4%, 64=0.1%, >=64=0.4%
>     issued rwts: total=10558,10990,0,0 short=0,0,0,0 dropped=0,0,0,0
>     latency   : target=0, window=0, percentile=100.00%, depth=256
> 
> Run status group 0 (all jobs):
>   READ: bw=10.4MiB/s (10.9MB/s), 10.4MiB/s-10.4MiB/s (10.9MB/s-10.9MB/s), io=1320MiB (1384MB), run=126934-126934msec
>  WRITE: bw=10.8MiB/s (11.3MB/s), 10.8MiB/s-10.8MiB/s (11.3MB/s-11.3MB/s), io=1374MiB (1440MB), run=126934-126934msec
> 
> Being unfamiliar with fio the interpretation is probably best left to others. 
> I'd expect a Pi2 to be much worse, no more than 4MB/s at best. 
> 

That was aarch64 code being tested. on aarch64 that supports
armv7, the armv7 code used via lib32 or chroot is mishandled
by FreeBSD and the test gets a bunch of failure notices. By
contrast, the same armv7 media booted on a real armv7 via
the armv7 kernel works fine.

I had been going to ask about your executing an armv7 build
of fio on the RPi4 with your drive, but for now it would not
provide the intended information.

I posted to the arm and current lists and submitted:

https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=290962

about the issue. I do not have 15.* or 14.* contexts to
test.

===
Mark Millard
marklmi at yahoo.com