(UEFI) Dell XPS 15z L511z modded BIOS - and HOWTO

[Florin9doi]

they have different aspect ratios

[jkbuha]

Yes they both have different aspect ratios - the problem is that there doesn't seem to be a standard above 1080p, especially on 15.6" screens.

Also I have another issue - in theory I could patch in the new resolution into the vbios for the new screen, but I'm worried that the connector may not support the increased bandwidth. Probably best I stick with a new 1080p screen for now until the resolution (and corresponding vbioses) become standardised above 1080p.

On the other topic re: Clover & Bios FSB - if anyone has any additional input over what kasar & I have discussed please let us know. Right now it seems like Clover+UEFI+FSB setting seems to be the best rational bet as we can easily get 25% increase in performance, but we need more testing on different processors for a detailed analysis. Thanks in advance!

[TimeWalker]

Oh you lucky people with dual channel connectors, my Vostro has a single channel connector and the only display I can install is the one I have - 1366x768, mat 14" ...

As for Clover - I will do some testing on FSB. Also, for some reason my laptop always freezes upon shutdown from OSX, not always really, but fairly frequently. UEFI drivers seem to cause this on my main install, but there are virtually no freezes with a clean install. Could you @jkbuha post your set of drivers, please?

[jkbuha]

Sure it's the standard driver set off the latest clover build (v953 i think). I have the same shutdown freezes btw.

[TimeWalker]

Thanks a bunch.

Post about it on projectosx please, in uefi thread, so that @dmazar be aware that I'm not the only one who's making this up

[jkbuha]

with pleasure - can you pm me the url?

[TimeWalker]

I will just post here so everyone having the issue could feedback.
http://www.projectosx.com/forum/index.ph...&start=960

[jkbuha]

Hmmm i don't have permission to reply to that topic. I dont really have time now to post elsewhere to gain points/reps/posts/whatever, so I'll wait until I have some bandwidth available...

[TimeWalker]

Oh yes, you have to pass *a quiz* for knowledge on basic hackintosh stuff to be able to post, sorry .. I completely forgot about this.
http://projectosx.com/forum/index.php?autocom=quiz

---

hah .. attempted to boot Legacy Windows thru Clover today.. I absolutely love people who coded error messages on this thing
first this:

then this:

and finally this ...

No way of booting it thru PBR either.. so only UEFI is an option on this crippled thing.

[TimeWalker]

the culprit of system freezing completely upon pressing Fn+F2 is the following:

Code:

Method (_Q8C, 0, NotSerialized)
                    {
                        P8XH (Zero, 0x8C)
                        Store (Zero, Local0)
                        Store (Zero, Local1)
                        Store (Zero, Local2)
                        Store (Zero, Local3)
                        Store (0xF2, WCA1)
                        //FOWM (0x11, 0x0B)
                        Store (WFI1, Local0)
                        Store (WFI2, Local1)
                        Store (0x0100, Local2)
                        Store (One, Local3)
                        If (WIFS)
                        {
                            Store (0x06, ^^^^AMW0.INF0)
                            Store (Zero, ^^^^AMW0.INF1)
                            Store (0xE008, ^^^^AMW0.INF2)
                            Store (Local0, ^^^^AMW0.INF3)
                            Store (Local1, ^^^^AMW0.INF4)
                            Store (Local2, ^^^^AMW0.INF5)
                            Store (Local3, ^^^^AMW0.INF6)
                            If (LEqual (DMFG, Zero))
                            {
                                Notify (AMW0, 0xD0)
                            }
                        }
                    }

same thing here

Code:

Method (_Q9F, 0, NotSerialized)
                    {
                        P8XH (Zero, 0x9F)
                        Store (Zero, Local0)
                        Store (Zero, Local1)
                        Store (Zero, Local2)
                        Store (Zero, Local3)
                        If (LEqual (QCKS, Zero))
                        {
                            Store (0xF2, WCA1)
                            //FOWM (0x11, 0x0B)
                        }
                        Store (WFI1, Local0)
                        Store (WFI2, Local1)
                        Store (0x0100, Local2)
                        Store (One, Local3)
                        If (WIFS)
                        {
                            Store (0x06, ^^^^AMW0.INF0)
                            Store (Zero, ^^^^AMW0.INF1)
                            Store (0xE008, ^^^^AMW0.INF2)
                            Store (Local0, ^^^^AMW0.INF3)
                            Store (Local1, ^^^^AMW0.INF4)
                            Store (Local2, ^^^^AMW0.INF5)
                            Store (Local3, ^^^^AMW0.INF6)
                            If (LEqual (DMFG, Zero))
                            {
                                Notify (AMW0, 0xD0)
                            }
                        }
                    }

If you comment out the call for method FOWM () it won't freeze, but won't work properly either. This exact call FOWM (0x11, 0x0B) appears to be causing the freeze for both of the queries it is used in. The only other query which call for this methods uses (0x11, 0x05) instead , this query is used for notifying the system about BIOS errors (based on comments from linux kernel driver as well as reading ACPI code).

The method itself is as follows, as you can see it takes 2 arguments :

Code:

Method (FOWM, 2, NotSerialized)
        {
            Acquire (PSMX, 0xFFFF) //
            Store (0xE0, WCMD)
            Store (0x57, TPID)
            Store (Arg0, WCCL)
            Store (Arg1, WCSE)
            Store (Zero, PHXI)
            Release (PSMX)
        }

I have no idea what PSMX is, must be something similar to MUTEX (or MUT0 how they are called in various asl codes). And furthermore I have no idea what are all of these registers meant for, but I guess some of them fails to allow write to it and system freezes. Will try writing some other value, 0x05 for example to see it it changes anything ...

UPDATE:

Made it call for FOWM (0x11, 0x05) - doesn't freeze, IOWMIFamily picks up the scancode properly: kernel[0]: WMIHIKeyboardDevice: Unknown key 0e08 i=7
Meaning the combination works, but won't do anything because system doesn't know how to parse it.. I could have coded this in, but the GUIDs are absent from DSDT and only GUID able to be notified is the one used to identify Dell WMI device, which is always notified by 0xD0. So I have no idea how to code the RF Kill behavior in... even though linux does it flawlessly.

UPDATE2:
Just attempted to write 0x0B into register WCSE from a brightness adjustment query, needles to say it froze. So something's is wrong there.. most likely the memory allocation is little bit off in UEFI mode.
This is A04:

Code:

OperationRegion (PNVS, SystemMemory, 0xBAE0D018, 0x100E)
        Field (PNVS, AnyAcc, NoLock, Preserve)
        {
            WCMD,   8,
            TPID,   8,
            WDID,   32,
            DI00,   32768
        }
        Field (PNVS, AnyAcc, NoLock, Preserve)
        {
            AccessAs (ByteAcc, 0x00),
            Offset (0x06),
            WCCL,   16,
            WCSE,   16,
            WCA1,   32,
            WCA2,   32,
            WCA3,   32,
            WCA4,   32,
            WCR1,   32,
            WCR2,   32,
            WCR3,   32,
            WCR4,   32
        }

This is A13:

Code:

OperationRegion (PNVS, SystemMemory, 0xBAE0C018, 0x100E)
        Field (PNVS, AnyAcc, NoLock, Preserve)
        {
            WCMD,   8,
            TPID,   8,
            WDID,   32,
            DI00,   32768
        }
        Field (PNVS, AnyAcc, NoLock, Preserve)
        {
            AccessAs (ByteAcc, 0x00),
            Offset (0x06),
            WCCL,   16,
            WCSE,   16,
            WCA1,   32,
            WCA2,   32,
            WCA3,   32,
            WCA4,   32,
            WCR1,   32,
            WCR2,   32,
            WCR3,   32,
            WCR4,   32
        }

As you can see it has changed from 0xBAE0D018 to 0xBAE0C018 across versions, but the bug still persists. Maybe we can find similar looking ACPI definitions from other OEMs?