[patch] move ich_init_opcodes() call back to where it was before r1461

[Jonathan A. Kollasch]

ich_init_opcodes() needs to be called after ichspi_lock is set.

Signed-off-by: Jonathan A. Kollasch

[Stefan Tauner]

On Sat, 12 Nov 2011 23:43:40 +0000
"Jonathan A. Kollasch"  wrote:
> ich_init_opcodes() needs to be called after ichspi_lock is set.
>
> Signed-off-by: Jonathan A. Kollasch

thanks jonathan for that patch, but this wont work as it should due to
ich_missing_opcodes.
i have prepared the attached patch already and will commit it as soon as
carldani had time to discuss it.
it changes the output which might not what we want:
[…]
SPIBAR = 0xfed1c000 + 0x3800
0x04: 0xe008 (HSFS)
HSFS: FDONE=0, FCERR=0, AEL=0, BERASE=1, SCIP=0, FDOPSS=1, FDV=1, FLOCKDN=1
WARNING: SPI Configuration Lockdown activated.
Reading OPCODES... done
0x06: 0x3f00 (HSFC)
HSFC: FGO=0, FCYCLE=0, FDBC=63, SME=0
0x08: 0x00003000 (FADDR)
0x50: 0x00000a0b (FRAP)
[…]

the problem is "Reading OPCODES... done" which is quite unexpected in
the middle of the register outputs...(?)

another option would be to make ichspi_lock a function or macro.
performance is probably not an issue(?) and we would avoid such ordering
problems.

[Carl-Daniel Hailfinger]

Am 13.11.2011 01:27 schrieb Stefan Tauner:
> On Sat, 12 Nov 2011 23:43:40 +0000
> "Jonathan A. Kollasch"  wrote:
> > > ich_init_opcodes() needs to be called after ichspi_lock is set.
> > >
> > > Signed-off-by: Jonathan A. Kollasch
>
> thanks jonathan for that patch, but this wont work as it should due to
> ich_missing_opcodes.
> i have prepared the attached patch already and will commit it as soon as
> carldani had time to discuss it.
> it changes the output which might not what we want:
> […]
> SPIBAR = 0xfed1c000 + 0x3800
> 0x04: 0xe008 (HSFS)
> HSFS: FDONE=0, FCERR=0, AEL=0, BERASE=1, SCIP=0, FDOPSS=1, FDV=1, FLOCKDN=1
> WARNING: SPI Configuration Lockdown activated.
> Reading OPCODES... done
> 0x06: 0x3f00 (HSFC)
> HSFC: FGO=0, FCYCLE=0, FDBC=63, SME=0
> 0x08: 0x00003000 (FADDR)
> 0x50: 0x00000a0b (FRAP)
> […]
>
> the problem is "Reading OPCODES... done" which is quite unexpected in
> the middle of the register outputs...(?)
>
> another option would be to make ichspi_lock a function or macro.
> performance is probably not an issue(?) and we would avoid such ordering
> problems.
>
> From: Stefan Tauner 
> Date: Sat, 12 Nov 2011 22:44:26 +0100
> Subject: [PATCH] ichspi: fix ich_init_opcodes() calls in ich_init_spi()
>
> By calling it early ichspi_lock was not set up correctly in accordance
> with the corresponding register, hence ich_init_opcodes() was always
> trying to programming the opcodes instead of reading them in from the
> opmenu in case of a locked down configuration.
>
> Thanks to Jonathan Kollasch for reporting this bug.

You forgot the A. in his name.
> Signed-off-by: Stefan Tauner

The cosmetic change in the output is not an issue for me.

Acked-by: Carl-Daniel Hailfinger

Regards,
Carl-Daniel

[Stefan Tauner]

On Sun, 13 Nov 2011 16:12:32 +0100
Carl-Daniel Hailfinger  wrote:
> Am 13.11.2011 01:27 schrieb Stefan Tauner:
> > Thanks to Jonathan Kollasch for reporting this bug.
>
> You forgot the A. in his name.

fixed :)
> The cosmetic change in the output is not an issue for me.
>
> Acked-by: Carl-Daniel Hailfinger

thanks, committed in r1464