Re: [coreboot] Mistery of the function: static void enable_rom_caching(void)

[Zoran Stojsavljevic]

Hello Rizwan,

Thank you for the tip. I am looking into the whole x86 asm file pointed by
you, and cranking it reading x86 architectural manual (lot of info to
digest). ;-)

If I'll have questions, I might come back later to ask for clarity, if you
allow me. :-)

Thank you again,
Zoran

On Mon, Aug 8, 2016 at 12:35 PM, Qureshi, Rizwan 
wrote:
> Hi Zoran,
>             For skylake refer src/soc/intel/skylake/bootblock/cache_as_ram.S
>
> around line 256. We are marking the ROMregion as WP and cacheable and
> programming an MTRR appropriately.
>
> So, it’s not removed buts its being done while setting up Cache As Ram.
>
> Regards,
>
> Rizwan Qureshi
>
> *From:* Zoran Stojsavljevic [mailto:[e-mail]
> ]
> *Sent:* Monday, August 8, 2016 12:33 PM
> *To:* coreboot ; Paul Menzel <paulepanter@users.
> sourceforge.net>
> *Cc:* Qureshi, Rizwan ; Ch, Naveenkrishna <
> [e-mail]>; Yang, York 
> *Subject:* Mistery of the function: static void enable_rom_caching(void)
>
> Hello community,
>
> I was/got again today in Coreboot directory src/soc/intel/braswell/bootblock/bootblock.c
> looking at static void enable_rom_caching(void).
>
> Then, since Paul Menzel have announcement ([ANNOUNCEMENT] Support for
> Intel Kaby Lake), I went to look into src/soc/intel/skylake/bootblock/bootblock.c,
> but I did not find this function. Kaby Lake patches do not support this
> function as well.
>
> Then I did the following on my WIN 10 (using VMware workstation 12) VM
> Fedora 24:
>
> [zoran@localhost intel]$ pwd
>
> /home/zoran/projects/coreboot/coreboot/src/soc/intel
>
> [zoran@localhost intel]$ git describe
>
> 4.4-1038-gdd65ef8
>
> [zoran@localhost intel]$ ls -al
>
> total 48
>
> drwxrwxr-x. 12 4096 Aug  1 10:02 .
>
> drwxrwxr-x. 14 4096 Aug  1 10:02 ..
>
> drwxrwxr-x.  5 4096 Aug  1 10:02 apollolake
>
> drwxrwxr-x.  6 4096 Aug  1 10:02 baytrail
>
> drwxrwxr-x.  6 4096 Aug  1 10:02 braswell
>
> drwxrwxr-x.  6 4096 Aug  1 10:02 broadwell
>
> drwxrwxr-x.  3 4096 Aug  1 10:02 common
>
> drwxrwxr-x.  7 4096 Aug  1 10:02 fsp_baytrail
>
> drwxrwxr-x.  7 4096 Aug  1 10:02 fsp_broadwell_de
>
> drwxrwxr-x.  5 4096 Aug  1 10:02 quark
>
> drwxrwxr-x.  3 4096 Aug  1 10:02 sch
>
> drwxrwxr-x.  7 4096 Aug  1 10:02 skylake
>
> [zoran@localhost intel]$ grep -r enable_rom *
>
> baytrail/bootblock/bootblock.c:static void enable_rom_caching(void)
>
> baytrail/bootblock/bootblock.c:          enable_rom_caching();
>
> braswell/bootblock/bootblock.c:static void enable_rom_caching(void)
>
> braswell/bootblock/bootblock.c:         enable_rom_caching();
>
> broadwell/bootblock/cpu.c:static void enable_rom_caching(void)
>
> broadwell/bootblock/cpu.c:     enable_rom_caching();
>
> fsp_baytrail/bootblock/bootblock.c:static void enable_rom_caching(void)
>
> fsp_baytrail/bootblock/bootblock.c:   enable_rom_caching();
>
> [zoran@localhost intel]$
>
> We see here that from skylake (also does not exists in apollolake) this
> function is dropped.
>
> Any logical explanation for this mismatch?
>
> Thank you,
>
> Zoran