Re: 13.0-RELEASE Poor performance on Raspberry pi 3 B+

On 2021-Nov-26, at 11:04, Mark Millard <marklmi@yahoo.com> wrote:

> On 2021-Nov-26, at 10:35, Juan David Hurtado G <jdhurtado@orbiware.com> wrote:
> 
>> Thank you so much for your time you all.
>> 
>> New microSD arrived today. It is a SanDisk Ultra, see this on the Raspberry
>> pi 3B+:
>> 
>> ```
>> root@generic:~ # dmesg | grep -i mmc
>> mmc0: <MMC/SD bus> on sdhci_bcm0
>> mmcsd0: 32GB <SDHC SD32G 8.5 SN A0169EA9 MFG 04/2021 by 3 SD> at mmc0
>> 50.0MHz/4bit/65535-block
>> ```
>> 
>> 50 MHz !! What a relief.
>> 
>> Other OS like RaspiOS and NetBSD runs well on the other cards, not sure
>> why. I still don't understand why sandisk is the only brand that works for
>> FreeBSD apparently.
> 
> I have non-sandisk microsd cards that work fine --and other sandisk
> models of microsd cards that also work. Mostly I use non-ultra sandisk
> ones as it turns out.

Some of these are not intended as high performance for
the type of context and some may not perform as well
as others, even if intended is high performance. I'm
just listing ones that have historically seemed to
work okay in my use:

Transcend High Endurance microSDHC
Kingston microSDHC I Industrial
Samsung EVO+ microSDHC I
Samsung EVO PLUS microSDXC I
SanDisk Extreme microSDHC I (A1)
SanDisk Extreme microSDXC I (A2)
SanDisk Extreme PLUS microSDXC I (A1)
SanDisk Ultra microSDHC I
SanDisk Ultra microSDXC I
SanDisk Ultra microSDXC I (A1)

So far all of these have worked. None of them are
recent purchases. I've never observed a small
MHz figure for any of them, as far as I remember
anyway.

I'm not claiming RPi* use of the fastest modes some
of the miscrosd cards are capable of.

> But once I could boot via USB I no longer use microsd cards very
> often in the RPi*'s that can boot from USB without a microsd card
> involved.
> 
>> I'm willing to give more information if needed or even to open a ticket in
>> bugzilla if you find it useful.
> 
> It does not seem that my direction of activity is
> relevant if it is media-type specific for the behavior.
> So I'll not be experimenting further with any RPi4B's
> (or otherwise).
> 
> But if you make a bugzilla report, various things about
> your RPi* configurations that you have reported would be
> appropriate to include as context.
> 
>> Have a great weekend !!
>> 
>> PS: I've also bought the usb ttl cable if I can give any more information
>> using the other card via serial console.
> 
> There are various settings for getting more debug
> output on the serial console for the RPi* stages
> of things. Some go in the config.txt but one goes
> in the eeprom content. As stands it does not seem
> that such would have helped for the specific problem:
> The problem is later, after the RPi* eeprom and
> firmware stages.
> 
> You may never need such, but at least you are forwarned
> about the possibility if you think it might help
> something in the future.
> 




===
Mark Millard
marklmi at yahoo.com
( dsl-only.net went
away in early 2018-Mar)