On Wed, Jun 11, 2025 at 01:55:32PM +0200, Casey Connolly wrote:
> Hi Varadarajan,
>
> On 5/7/25 09:07, Varadarajan Narayanan wrote:
> > Using the 'reboot edl' command in Linux, the platform can reboot to the
> > Emergency Download mode. Implement the same for U-Boot.
> >
> > v3: * Introduce a sysreset op that will pass down the 'reset' command
> > arguments to registered handlers
> > * Handle 'reset to edl' alone in qcom-psci driver
> >
> > v2: * Rebased to recent sources
> > * Dropped the first patch as it is not applicable anymore
> > * Bind the new driver from psci driver
>
>
> Sorry for the late reply.
>
> This really needs some associated documentation, at least adding the edl arg
> to doc/usage/cmd/reset.rst and the help output in cmd/boot.c should describe
> the edl arg if CONFIG_SYSRESET_QCOM_PSCI is enabled.
ok.
>
> it also doesn't build?
>
> ../drivers/sysreset/sysreset-uclass.c: In function 'sysreset_walk_arg':
> ../drivers/sysreset/sysreset-uclass.c:93:63: error: passing argument 3 of
> 'sysreset_request_arg' from incompatible pointer type
> [-Wincompatible-pointer-types]
> 93 | ret = sysreset_request_arg(dev, argc, argv);
> | ^~~~
> | |
> | char * const*
> ../drivers/sysreset/sysreset-uclass.c:35:69: note: expected 'const char **'
> but argument is of type 'char * const*'
> 35 | int sysreset_request_arg(struct udevice *dev, int argc, const char
> *argv[])
ok.
> ignoring that for a sec, I'm a bit worried still about how this behaves on
> boards that don't implement this psci call, on sdm845 for example we just
> hang.
>
> I think psci_system_reset2() should be adjusted to time out (though I can't
> imagine a psci implementation that would return execution to el1 during the
> reset, so maybe the timeout isn't even needed?) and return an error if the
> reset didn't happen.
Would checking if the system-reset2 function is implemented in PSCI and
handling appropriately similar to [1] be ok? Please let me know.
> Then qcom_psci_sysreset_request_arg() should print some useful error message
> ("reset2 not supported on this platform, reset to edl failed") and bubble up
> the error code to abort the reset entirely.
If the function is not implemented will print a message and bubble up
-EPROTONOSUPPORT.
1 -
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/firmware/psci/psci.c#n554
Thanks
Varada
