GNUtoo has quit [Remote host closed the connection]
GNUtoo has joined #u-boot
GNUtoo has quit [Ping timeout: 268 seconds]
GNUtoo has joined #u-boot
mmu_man has joined #u-boot
persmule has quit [Remote host closed the connection]
persmule has joined #u-boot
apritzel has joined #u-boot
prabhakarlad has joined #u-boot
mmu_man has quit [Ping timeout: 268 seconds]
hanetzer1 is now known as hanetzer
persmule has quit [Remote host closed the connection]
persmule has joined #u-boot
davlefou has quit [Ping timeout: 268 seconds]
mmu_man has joined #u-boot
davlefou has joined #u-boot
monstr has joined #u-boot
<marex>
sjg1: well, I have nothing to sign since whatever I have now does no longer even produce the SPL with IVT/bootdata, so zero progress on signing obviously
frieder has joined #u-boot
<milkylainen>
Yeah. STM versions of TF-A/U-Boot just won't play nice.
<marex>
milkylainen: that's why you should only use mainline
<milkylainen>
marex: I know. But mainline TF-A doesn't support TBBR standing on STM hw keyhash.
torez has joined #u-boot
<marex>
milkylainen: you probably dont need ATF at all btw
<milkylainen>
marex: My customer thinks otherwise.
<milkylainen>
marex: How do you mean btw?
<milkylainen>
hw fsbl keying as spl and go?
<milkylainen>
u-boot spl
<marex>
yes, make HW authenticate SPL and go from there
GNUtoo has quit [Remote host closed the connection]
<hramrach>
that's actually a good question - when do you need TFA?
<hramrach>
it seems to provide PSCI at least on some platforms
GNUtoo has joined #u-boot
<marex>
hramrach: never ?
<marex>
hramrach: on aargh64 PSCI is mandated by spec
<marex>
on armv7 it is entirely optional
<marex>
whatever PSCI provider there is, it doesnt have to be atf
lucascastro has joined #u-boot
<hramrach>
yes, I suppose it's like if you do have ATF it's the logical PSCI provider, and it's code duplication to implement it elsewhere in case you did not use ATF so you need ATF on aargh64
persmule has quit [Remote host closed the connection]
persmule has joined #u-boot
<jkridner>
is there a simple way to make debug() statements in a file print out? #define LOG_DEBUG isn't enough on it's own.
mmu_man has quit [Ping timeout: 268 seconds]
wyre_ is now known as wyre
persmule has quit [Remote host closed the connection]
<hramrach>
#define DEBUG works for me
persmule has joined #u-boot
gsz has joined #u-boot
<hramrach>
and it probably needs to be defined before some includes
<marex>
before common.h , i.e. just stick it at the beginning of the file
<jkridner>
hramrach: I put it at the very top.
<jkridner>
does it go into a special log or just print out on the serial port?
<jkridner>
I've tried both #define DEBUG and #define LOG_DEBUG
<jkridner>
I've tried turning on logging.
<jkridner>
oddly, I haven't gotten the 'log' command there yet, despite adding it as well. The timestamp seems current.
<jkridner>
giving up, swapping debug w/ printf.
<jkridner>
ok, looks like my problem could simply be that my debug statements aren't being reached as the printf statements didn't print either!
<sjg1>
marex: What is the 'mkimage' command line used at present? It is quite unclear to me. We can replicate that in binman. Also I wonder what is the 0x7E1000 value in the .cfg file as we need to replicate that also?
prabhakarlad has quit [Quit: Client closed]
vagrantc has joined #u-boot
<milkylainen>
is the github release archive considered a main release archive?
<vagrantc>
unless ... something has changed, in which case i'd love to know :)
prabhakarlad has joined #u-boot
prabhakarlad has quit [Quit: Client closed]
persmule has quit [Read error: Connection reset by peer]
persmule has joined #u-boot
<Tartarus>
milkylainen No, the official releases are where vagrantc said just now, with signed tags and everything
<Tartarus>
for tarballs
<marex>
sjg1: err ... whatever is in upstream u-boot right now just works
<vagrantc>
yay! i'm not totally obsolete! :)
lucascastro has quit [Ping timeout: 268 seconds]
monstr has quit [Remote host closed the connection]
Xeroine has joined #u-boot
<sjg1>
Marex do you know the answers or can you find out?
prabhakarlad has joined #u-boot
<apritzel>
Tartarus: I am on the DM port of the SMC91111 ethernet driver right now (I know, I am late), what would be best way going forward? Apply your series, then send a revert plus the conversion?
<Tartarus>
apritzel: How close are you to being done?
<Tartarus>
I'm not going to apply it today, but I want to later this week
<apritzel>
I will probably finish the code this week, but it wouldn't be really ready at this point, I guess
<Tartarus>
Maybe i'll just partly apply the series this week, need to see what feedback if any I get from NXP
<Tartarus>
Given the platforms and functionality I'm dropping there.
<apritzel>
Tartarus: actually dropping some users of the smc91111 driver solved a big problem for me, as I now don't need to support those (non-DM, non-ofcontrol) anymore
<apritzel>
so can do a clean DM-only conversion
<Tartarus>
Ah, yeah, ok
<Tartarus>
If it's not OF_CONTROL it's likely abandoned
lucascastro has quit [Read error: Connection reset by peer]
camus has quit [Ping timeout: 268 seconds]
camus has joined #u-boot
frieder has quit [Remote host closed the connection]
<Tartarus>
In case anyone else is seeing the azure macOS brownout message, I do plan to grab the macOS/ubuntu-22.04 patches I posted probably later today
<Tartarus>
Testing some related azure work atm
akaWolf has quit [Ping timeout: 252 seconds]
akaWolf has joined #u-boot
apritzel has quit [Ping timeout: 268 seconds]
prabhakarlad has quit [Quit: Client closed]
prabhakarlad has joined #u-boot
vfazio has quit [Read error: Connection reset by peer]
vfazio has joined #u-boot
prabhakarlad has quit [Quit: Client closed]
torez has quit [Quit: torez]
akaWolf has quit [Ping timeout: 268 seconds]
<johang>
lvrp16: same problem with that build
akaWolf has joined #u-boot
umbramalison has quit [Remote host closed the connection]
umbramalison has joined #u-boot
<marex>
sjg1: I think I am still missing something obvious
<marex>
sjg1: shouldn't the binman pick the bundled SPL+DDR blobs and run mkimage on that already ?
sbach has quit [Read error: Connection reset by peer]
eternalforms has quit [Remote host closed the connection]
<marex>
Tartarus: it should ideally get as much testing as possible before release, in case there are more problems
<marex>
Peng_Fan: ^ seems to be assigned to you, are you now maintainering the mmc subsystem ? :)
Xeroine has quit [Ping timeout: 252 seconds]
<sjg1>
marex: You can actually see what it does if you look at mkimage.py - see the run_cmd() line
<sjg1>
marex: So mkimage -d <input-data-file> <args from the description> <output-file-name>
<lvrp16>
johang: can you pm me the script? I will give it a go.
<sjg1>
marex: args-from-description is "-n spl/u-boot-spl.cfgout -T imx8mimage -e 0x7e1000"
<sjg1>
marex: So actually the -n arg is used to pass in another file. That's a non-standard usage of mkimage, I think. Will need to update binman to support it
<sjg1>
marek: But my question was really that I cannot see where the 'mkimage
<sjg1>
marex: command is run
<marex>
sjg1: I am confused ...
<marex>
sjg1: you write "I cannot see where the 'mkimage..command is run" ... yet a few lines above "look at mkimage.py - see the run_cmd() line"
<sjg1>
marex: Not as confused as I am
<sjg1>
I mean I cannot see where it runs in mainline
<marex>
arch/arm/dts/imx8mm-u-boot.dtsi mkimage {} should trigger the mkimage which generates the SPL with IVT/boot-data section, no ?
<sjg1>
marex: Ah I see
<sjg1>
marex: So we need a way to tell binman to pass a filename in with the -n argument. Does any other board do that? If it is just imx then we can just do a one-off
<marex>
kirkwood does per git grep
<sjg1>
marex: Yes it seems several others use -n as a filename
<sjg1>
marex: Let me see how best to implement this
<marex>
btw it seems ELCE will be a ghost town
<marex>
very few people go
<sjg1>
marex: why not?
<marex>
various reasons
<sjg1>
Oh well, maybe next year
<marex>
sjg1: do you expect things to get any better ?
<sjg1>
marex: with what? ELCE?
<marex>
well that, covid, conferences ...
<sjg1>
Well osfc is going ahead and is normal, so far as I know. Maybe that's why it is in Sweden? I heard that they gave up on lockdowns long before the UK, etc
<sjg1>
marex: just getting back to this...the u-boot-spl-ddr.bin is passed to mkimage twice: once as the input filename and another as the -n argument (via the .cfg file). Why?
akaWolf has quit [Ping timeout: 252 seconds]
<marex>
sjg1: I don't know, although I suspect the -n cfg file is the relevant one
<marex>
since that's what tells mkimage imx8mimage where to place that SPL in DRAM
akaWolf has joined #u-boot
<Forty-Bot>
sjg1: -n is standard AFAICT
<Forty-Bot>
it and -R are free-form for whatever the image type wants to do
<Forty-Bot>
I'd say round 2/3 of the image types take a file to -n
<Forty-Bot>
IMO it's fitimage which is unusual for using -f instead of -n