Re: [SeaBIOS] SeaBIOS recognising USB 3.0 on boot works - partly
On Sun, Dec 13, 2015 at 09:16:02PM +0000, edward wandasiewicz wrote:
On Sun, Dec 13, 2015 at 9:05 PM, edward wandasiewicz <0.w3ntd@gmail.com> wrote:
Hooray, it works... minus 2 scenarios :(
With both Philips USB and Mushkin USB attached, and I perform a cold boot
1. YES - USB 3 & USB 3 - once each (Philips & Mushkin)
cbmem.yes.both.USB3.USB3.gz
2. NO - USB 3 & USB 3 - once each (Philips & Muskin) - but double Philips showing
cbmem.yes.both.USB3.USB3.DoublePhilips.gz
3. YES - USB 3 & Type C - once each (Philips & Mushkin)
cbmem.yes.both.USB3.TypeC.gz
I can't get a double Philips to appear
4. YES - Type C & Type C - once each (Philips & Mushkin)
cbmem.yes.both.TypeC.TypeC.gz
5. NO - Type C & Type C - one Philiips & no Mushkin
I forgot to save
$ cbmem -c
and now can't replicate it. :(
If you can't reproduce, it may have been a transient hardware issue (eg, loose plug). Otherwise, to summarize, the devices are now detected pretty reliably, but you still have an issue with double detection of the Philips drive? The double detection looks like the controller doing something weird. I need to look closer at the spec, but I don't think I'll be able to do that until later in the week.
When I do, I will post output.
When attempting a cold boot - shutdown - cold boot - shutdown - cold boot - shutdown - ....
It seems the list of bootable devices can change order, yet the devices are plugged into the same ports all the time, and never unplugged.
By default, the order of USB drives in the menu is basically random. If you want to ensure a particular order for USB then a "bootorder" file can be created. You can look at the wiki for info on how to do that, or you can try to extract the bootorder file that John Lewis uses in his roms. -Kevin
On 12/14/2015 6:02 PM, Kevin O'Connor wrote:
On Sun, Dec 13, 2015 at 09:16:02PM +0000, edward wandasiewicz wrote:
On Sun, Dec 13, 2015 at 9:05 PM, edward wandasiewicz <0.w3ntd@gmail.com> wrote:
When I do, I will post output. When attempting a cold boot - shutdown - cold boot - shutdown - cold boot - shutdown - ....
It seems the list of bootable devices can change order, yet the devices are plugged into the same ports all the time, and never unplugged. By default, the order of USB drives in the menu is basically random. If you want to ensure a particular order for USB then a "bootorder" file can be created. You can look at the wiki for info on how to do that, or you can try to extract the bootorder file that John Lewis uses in his roms.
-Kevin
bootorder files with preference for USB and SSD can also be found in my github, no extraction necessary :) https://github.com/MattDevo/coreboot/tree/master/cbfs
On Mon, Dec 14, 2015 at 07:02:14PM -0500, Kevin O'Connor wrote:
The double detection looks like the controller doing something weird. I need to look closer at the spec, but I don't think I'll be able to do that until later in the week.
After reviewing the spec I have a guess to what is occurring. Can you pull the "testing" branch again and retry? -Kevin
I managed to replicate a scenario 2 but with Type-C instead of USB 3, with both 42-g9c58583 and 43-g55de21d 2(b) NO - Type C & Type C - once each (Philips & Muskin) - but double Philips showing cbmem.yes.both.TypeC.TypeC.DoublePhilips.42-g9c58583 cbmem.yes.both.TypeC.TypeC.DoublePhilips.43-g55de21d I thought, lets try swapping over the devices, but keeping the same Type-C cables on the left and right hand side and see what happens. I get scenario 5. 5. NO - Type C & Type C - one Philiips & no Mushkin cbmem.no.TypeC.TypeC.one.Philips.no.Mushkin What I did notice, was there was no "yellow light" appear on the Mushkin device, when I saw the SeaBIOS list of bootable devices. It was only after moving the cable, or letting a few seconds pass (5-10+), or a combination of both, did the yellow light on the Mushkin device start to appear, indicating the device was receiving power. What I then noticed, was when booting the Pixel 2015, and the Chrome "OS Verification is OFF" is displayed, the Mushkin "yellow light" appears, indicating the device is receiving power, before we ever get to SeaBIOS, which makes me think scenario 5 is a "device not receiving power properly" for some reason. I'm starting to wish I never found either of these 2 USB devices. I'm sure we'll get to the bottom of it... :) Edward. On Thu, Dec 17, 2015 at 10:17 PM, Kevin O'Connor <kevin@koconnor.net> wrote:
On Mon, Dec 14, 2015 at 07:02:14PM -0500, Kevin O'Connor wrote:
The double detection looks like the controller doing something weird. I need to look closer at the spec, but I don't think I'll be able to do that until later in the week.
After reviewing the spec I have a guess to what is occurring. Can you pull the "testing" branch again and retry?
-Kevin
participants (3)
-
edward wandasiewicz -
Kevin O'Connor -
Matt DeVillier