lk/linux-kernel-test
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

arch/tile: parameterize system PLs to support KVM port

Browse Source 
While not a port to KVM (yet), this change modifies the kernel
to be able to build either at PL1 or at PL2 with a suitable
config switch.  Pushing up this change avoids handling branch
merge issues going forward with the KVM work.

Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
This commit is contained in:
Chris Metcalf
2010-10-14 16:23:03 -04:00
parent bf65e440e8
commit a78c942df6
23 changed files with 337 additions and 131 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
arch/tile/mm/fault.c
View File

@ -563,10 +563,10 @@ do_sigbus:
/*
* When we take an ITLB or DTLB fault or access violation in the
* supervisor while the critical section bit is set, the hypervisor is
* reluctant to write new values into the EX_CONTEXT_1_x registers,
* reluctant to write new values into the EX_CONTEXT_K_x registers,
* since that might indicate we have not yet squirreled the SPR
* contents away and can thus safely take a recursive interrupt.
* Accordingly, the hypervisor passes us the PC via SYSTEM_SAVE_1_2.
* Accordingly, the hypervisor passes us the PC via SYSTEM_SAVE_K_2.
*
* Note that this routine is called before homecache_tlb_defer_enter(),
* which means that we can properly unlock any atomics that might
@ -610,7 +610,7 @@ struct intvec_state do_page_fault_ics(struct pt_regs *regs, int fault_num,
* fault. We didn't set up a kernel stack on initial entry to
* sys_cmpxchg, but instead had one set up by the fault, which
* (because sys_cmpxchg never releases ICS) came to us via the
* SYSTEM_SAVE_1_2 mechanism, and thus EX_CONTEXT_1_[01] are
* SYSTEM_SAVE_K_2 mechanism, and thus EX_CONTEXT_K_[01] are
* still referencing the original user code. We release the
* atomic lock and rewrite pt_regs so that it appears that we
* came from user-space directly, and after we finish the

2
arch/tile/mm/init.c
View File

@ -1060,7 +1060,7 @@ void free_initmem(void)
/*
* Free the pages mapped from 0xc0000000 that correspond to code
* pages from 0xfd000000 that we won't use again after init.
* pages from MEM_SV_INTRPT that we won't use again after init.
*/
free_init_pages("unused kernel text",
(unsigned long)_sinittext - text_delta,