ChanServ changed the topic of #glasgow to: glasgow debug tool · code https://github.com/GlasgowEmbedded/glasgow · forum https://glasgow.whitequark.org · logs https://freenode.irclog.whitequark.org/glasgow · production https://www.crowdsupply.com/1bitsquared/glasgow · no ETAs at the moment
bvernoux has quit [Quit: Leaving]
<d1b2> <esden> Streaming tomorrow again: https://twitter.com/esden/status/1265091803214045185?s=20
<d1b2> <esden> ||There might be some... "esden gives into peer pressure" moment ...||
<d1b2> <esden> heh interesting how the bridge essentially ignores post edits that I do on the discord side... interesting
<d1b2> <esden> ahh well... it is what it is 😄
<d1b2> <edbordin> I was also wondering if the discord users could develop a secret code using reaction emojis and embedded images 😛
<d1b2> <esden> yeah reaction emoji don't translate. Embedded pictures should be posted as links to IRC... not sure.
<d1b2> <edbordin> you're right, it does paste a link in IRC
<d1b2> <esden> yeah so that works fine then 🙂
<d1b2> <esden> On another note. I added the slashes to the pullup pulldown legends... I also mention HiZ...
<whitequark> esden: I think hiz is a bit misleading there
<whitequark> since the pin isn't necessarily hiz if the gpio expander is inactive
<whitequark> i.e. you can't tristate the pin using the expander
<d1b2> <esden> Yeah, right... so how would you call the third mode you set the expanders to?
<d1b2> <esden> @tnt complained that just saying Pull-Up/-Down is not enough... :/
<d1b2> <Darius> passive?
<whitequark> i'd say PU/PD
<whitequark> PU/PD/P- seems fine too
<whitequark> (- could mean "negative" but it's unambiguous in context I think?)
<whitequark> somethin galong those lines, anyway
<d1b2> <esden> Added labels indicating which shifter is responsible for which signal. Might be useful when probing around the board.
<d1b2> <Darius> it's not like there is room for an essay, a reminder is enough - RTFM if you aren't sure.. 🙂
<d1b2> <esden> @Darius yeah, indeed, I am just trying to do some useful "memory support" stuff on the board, and when I show the board to people at events, so they get some clues of what it does.
<d1b2> <Darius> yeah, perhaps print up a larger reference card as well?
<d1b2> <esden> even if it is not immediately clear what PU/PD shortcut means for some people
<d1b2> <esden> Yeah I definitely will try to make a card like I did for the 1Bitsy
<d1b2> <esden> or iCEBreaker
<d1b2> <Darius> yeah they are very cool
<d1b2> <esden> (iCEBreaker still does not have a hardcopy card >_<)
<d1b2> <esden> but the legend for iCEBreaker is pretty big... humm... I need to figure it out 😄
<d1b2> <Darius> popup card 😉
<d1b2> <esden> "let me just leave this 10kg reference book here"
<d1b2> <tnt> @esden Huh ? No, I just didn't associate PUPD as meaning "Pull Up / Pull Down". I read it as "Pull UP Data" or something like that ...
<d1b2> <Rogan> Crazy question: I saw Armbian was showing off a board that switches 8 UART’s to one, and 8 SD cards between SBC’s under test, and the test controller. The idea is to allow a controller to write the boot image to SD, then switch it over to the DUT, apply power and monitor boot logs. Went down a bit of a rabbit hole looking for SD card emulators, where every conclusion was “we’d need an fpga/cpld to do this properly”. Any thoughts of using
<d1b2> Glasgow as an SD card emulator?
<whitequark> not gonna happen
<whitequark> well
<whitequark> wait
<whitequark> let me refresh my memory of the SD protocol
<d1b2> <Rogan> It’s SPI based
<d1b2> <Rogan> Afaik
<d1b2> <Rogan> Up to qspi
<whitequark> I think it's not really SPI based, which is actually good for our case
<whitequark> so if I recall correctly you can actually have the card respond to the SD host indefinitely with wait states
<whitequark> this is just what Glasgow would need to have processing on the host
<whitequark> the USB host that drives the Glasgow
<d1b2> <Rogan> Right
<whitequark> in SD mode, anyway
<whitequark> in SPI mode I don't recall how it works
<whitequark> I think it still has wait states but they work differently
<d1b2> <eddyb> @esden heh, I like the slashes, but I think you missed the label on the B resistors (vs the A ones)
<d1b2> <Darius> @Rogan I pondered a smaller version of that for work which would let you switch an SD card between an SBC and a USB SD card reader plus ancilliary reset logic etc
<d1b2> <Darius> never got around to it in the end, although I did get some micro SD card extenders so you didn't have to open the device when you want to reflash it..
<d1b2> <Rogan> This was the tweet that started me down this path: https://twitter.com/armbian/status/1264930829811232769?s=21
Stormwind_mobile has quit [Ping timeout: 258 seconds]
<d1b2> <Rogan> Looks neat, and probably better in fact for bulk operations, since it would scale higher than a Glasgow could. But for a single device, under heavy development, Glasgow could offer a big speedup, paired with suitable adapters, and a power control option.
<d1b2> <Darius> probably would have bought that if it existed when I was looking!
<d1b2> <Rogan> Apparently ZipCPU has an implementation in Verilog: https://github.com/ZipCPU/sdspi
<d1b2> <Mathis> Is there an I2C version for @esden ?
Stormwind_mobile has joined #glasgow
<d1b2> <Hedge> I need to finish test and design of the cost-down version of spispy (ice40 + PSRAM SPINOR emulator)
<d1b2> <Hedge> Would be fun to make a big brother of it for SD Cards
<d1b2> <Hedge> Maybe ECP5 + DDR3 SODIMM socket?
<d1b2> <Rogan> 8GB SODIMM == 8GB SD card? Sounds reasonable.
<d1b2> <daveshah> There is no latency constraint on SD cards, right?
<d1b2> <Rogan> Wonder what the largest SODIMM available would be...
<d1b2> <daveshah> unlike emulating other memory devices
<d1b2> <Hedge> @daveshah good question
<d1b2> <Hedge> I haven’t looked at the specs in a long time
<d1b2> <daveshah> Given they are manage dNAND based there must be some latency flexibility as random reads wouldn't be totally instanteous
<d1b2> <Hedge> And there are some pretty terrible ones, speed wise
<d1b2> <tnt> Yeah as wq mentionned earlier you "poll" the SD until it's ready.
<d1b2> <Hedge> Ah
<d1b2> <tnt> as least in some version of the protocol ... SD has so many modes ...
<d1b2> <Hedge> @Rogan max appears to be 16GB per SODIMM (at least by what’s available)
<d1b2> <Hedge> @tnt that’s much more forgiving than SPINOR
<d1b2> <Rogan> 16GB should be enough for anyone
<d1b2> <Rogan> 😄
<d1b2> <Rogan> Feels like that got a little offtrack, though. The original question was whether emulating an SD card could be done with just a Glasgow, adding a SODIMM socket to a carrier board seems like going a bit too far from that.
<d1b2> <Hedge> Seems par for the course 😉
ec0 has joined #glasgow
pg12_ has quit [Quit: pg12_]
pg12 has joined #glasgow
pg12 has quit [Client Quit]
pg12 has joined #glasgow
pg12 has quit [Client Quit]
pg12 has joined #glasgow
<d1b2> <TomKeddie> @Darius I tried that once, made a mess of all the different power supply domains and all the different parts were powering each other through their protection diodes. A learning experience. The sd card needed to be power cycled to switch it between spi and sd modesbas I recall.
josi9824 has joined #glasgow
josi9824 is now known as josi
ali_as has quit [Remote host closed the connection]
ali_as has joined #glasgow
samlittlewood has quit [Quit: samlittlewood]
emilazy has quit [Ping timeout: 272 seconds]
levi has quit [Ping timeout: 260 seconds]
emilazy has joined #glasgow
emilazy has quit [Ping timeout: 272 seconds]
analprolapse has quit [Ping timeout: 260 seconds]
levi has joined #glasgow
emilazy has joined #glasgow
analprolapse has joined #glasgow
Stormwind_mobile has quit [Ping timeout: 260 seconds]
Stormwind_mobile has joined #glasgow
levi has quit [Quit: Connection closed for inactivity]
bvernoux has joined #glasgow
Stormwind_mobile has quit [Ping timeout: 258 seconds]
Stormwind_mobile has joined #glasgow
bvernoux has quit [Quit: Leaving]
Stormwind_mobile has quit [Read error: Connection reset by peer]
Stormwind_mobile has joined #glasgow
<gruetzkopf> Go with -c
<whitequark> -c?
<gruetzkopf> argh, wanted to hit the twitch chat
<electronic_eel> the twitch chat and irc seems not to be connected
<gruetzkopf> it is not
jn__ has quit [Ping timeout: 272 seconds]
jn__ has joined #glasgow
ali_as has quit [Remote host closed the connection]
ali_as has joined #glasgow
<d1b2> <esden> yeah sorry, It usually is a bit annoying to those that don't watch the streams, if you dump all the out of context stuff into the main chat channel. That is why I decided to just use the twitch chat as the on stream chat.
<d1b2> <esden> Ohh @0xdec thank you for sending me the usb-c layout recommendation: https://microchipsupport.force.com/s/article/USB-Type-C-Layout-Recommendations
<d1b2> <esden> That does have some interesting tips.
<d1b2> <esden> The result after the stream. Definitely more polish and poking required. But I think it is a good first attempt. 🙂
<whitequark> very nice
<d1b2> <esden> the usb-c 2.0 connectors are ... let's say... "tricky"
<d1b2> <esden> @whitequark thank you! 🙂
<d1b2> <esden> ohh wow I just see something
<d1b2> <esden> the U15 can be rotated 90 ccw and we can connect to the A9 pad for VBUS
<d1b2> <esden> I tink the diff pair will have a more straight path that way
<d1b2> <0xdec> @esden good idea!
Stormwind_mobile has quit [Read error: Connection reset by peer]
Stormwind_mobile has joined #glasgow
<d1b2> <Darius> @TomKeddie one of those harder-than-it-looks deals I suppose