diff options
author | Chris Metcalf <cmetcalf@tilera.com> | 2013-08-10 12:35:02 -0400 |
---|---|---|
committer | Chris Metcalf <cmetcalf@tilera.com> | 2013-08-30 11:56:58 -0400 |
commit | 35f059761c5ac313d13372fe3cdaa41bce3d0dbf (patch) | |
tree | 1a8f7e0eba01afac74c081348530fccd63dc48e4 /arch/tile/kernel/intvec_64.S | |
parent | 4036c7d3542ce82ea343bf95dd05ca46aefba9aa (diff) | |
download | linux-35f059761c5ac313d13372fe3cdaa41bce3d0dbf.tar.bz2 |
tilegx: change how we find the kernel stack
Previously, we used a special-purpose register (SPR_SYSTEM_SAVE_K_0)
to hold the CPU number and the top of the current kernel stack
by using the low bits to hold the CPU number, and using the high
bits to hold the address of the page just above where we'd want
the kernel stack to be. That way we could initialize a new SP
when first entering the kernel by just masking the SPR value and
subtracting a couple of words.
However, it's actually more useful to be able to place an arbitrary
kernel-top value in the SPR. This allows us to create a new stack
context (e.g. for virtualization) with an arbitrary top-of-stack VA.
To make this work, we now store the CPU number in the high bits,
above the highest legal VA bit (42 bits in the current tilegx
microarchitecture). The full 42 bits are thus available to store the
top of stack value. Getting the current cpu (a relatively common
operation) is still fast; it's now a shift rather than a mask.
We make this change only for tilegx, since tilepro has too few SPR
bits to do this, and we don't need this support on tilepro anyway.
Signed-off-by: Chris Metcalf <cmetcalf@tilera.com>
Diffstat (limited to 'arch/tile/kernel/intvec_64.S')
-rw-r--r-- | arch/tile/kernel/intvec_64.S | 21 |
1 files changed, 10 insertions, 11 deletions
diff --git a/arch/tile/kernel/intvec_64.S b/arch/tile/kernel/intvec_64.S index df19d4f3946e..3b35bb490d3e 100644 --- a/arch/tile/kernel/intvec_64.S +++ b/arch/tile/kernel/intvec_64.S @@ -132,13 +132,9 @@ intvec_\vecname: mfspr r3, SPR_SYSTEM_SAVE_K_0 /* Get &thread_info->unalign_jit_tmp[0] in r3. */ + bfexts r3, r3, 0, CPU_SHIFT-1 mm r3, zero, LOG2_THREAD_SIZE, 63 -#if THREAD_SIZE < 65536 - addli r3, r3, -(PAGE_SIZE - THREAD_INFO_UNALIGN_JIT_TMP_OFFSET) -#else - addli r3, r3, -(PAGE_SIZE/2) - addli r3, r3, -(PAGE_SIZE/2 - THREAD_INFO_UNALIGN_JIT_TMP_OFFSET) -#endif + addli r3, r3, THREAD_INFO_UNALIGN_JIT_TMP_OFFSET /* * Save r0, r1, r2 into thread_info array r3 points to @@ -365,13 +361,13 @@ intvec_\vecname: 2: /* - * SYSTEM_SAVE_K_0 holds the cpu number in the low bits, and - * the current stack top in the higher bits. So we recover - * our stack top by just masking off the low bits, then + * SYSTEM_SAVE_K_0 holds the cpu number in the high bits, and + * the current stack top in the lower bits. So we recover + * our starting stack value by sign-extending the low bits, then * point sp at the top aligned address on the actual stack page. */ mfspr r0, SPR_SYSTEM_SAVE_K_0 - mm r0, zero, LOG2_THREAD_SIZE, 63 + bfexts r0, r0, 0, CPU_SHIFT-1 0: /* @@ -393,6 +389,9 @@ intvec_\vecname: * cache line 1: r6...r13 * cache line 0: 2 x frame, r0..r5 */ +#if STACK_TOP_DELTA != 64 +#error STACK_TOP_DELTA must be 64 for assumptions here and in task_pt_regs() +#endif andi r0, r0, -64 /* @@ -690,7 +689,7 @@ intvec_\vecname: } { shl16insli r21, r21, hw1(__per_cpu_offset) - bfextu r20, r20, 0, LOG2_THREAD_SIZE-1 + bfextu r20, r20, CPU_SHIFT, 63 } shl16insli r21, r21, hw0(__per_cpu_offset) shl3add r20, r20, r21 |