powerpc/math-emu: Allow math-emu to be used for HW FPU

(Including 64-bit ones)

This allow SW emulation by the kernel of optional instructions
such as fsqrt which aren't implemented on some processors, and
thus fixes some Fedora 19 issues such as Anaconda since the
compiler is set to generate those by default on 64-bit.

Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
This commit is contained in:
Benjamin Herrenschmidt 2013-06-09 17:01:24 +10:00
parent 04ae900171
commit 4e63f8edfe
2 changed files with 16 additions and 2 deletions

View file

@ -298,7 +298,7 @@ config HUGETLB_PAGE_SIZE_VARIABLE
config MATH_EMULATION
bool "Math emulation"
depends on 4xx || 8xx || E200 || PPC_MPC832x || E500
depends on 4xx || 8xx || PPC_MPC832x || BOOKE
---help---
Some PowerPC chips designed for embedded applications do not have
a floating-point unit and therefore do not implement the
@ -307,6 +307,10 @@ config MATH_EMULATION
unit, which will allow programs that use floating-point
instructions to run.
This is also useful to emulate missing (optional) instructions
such as fsqrt on cores that do have an FPU but do not implement
them (such as Freescale BookE).
config PPC_TRANSACTIONAL_MEM
bool "Transactional Memory support for POWERPC"
depends on PPC_BOOK3S_64

View file

@ -1125,7 +1125,17 @@ void __kprobes program_check_exception(struct pt_regs *regs)
* ESR_DST (!?) or 0. In the process of chasing this with the
* hardware people - not sure if it can happen on any illegal
* instruction or only on FP instructions, whether there is a
* pattern to occurrences etc. -dgibson 31/Mar/2003 */
* pattern to occurrences etc. -dgibson 31/Mar/2003
*/
/*
* If we support a HW FPU, we need to ensure the FP state
* if flushed into the thread_struct before attempting
* emulation
*/
#ifdef CONFIG_PPC_FPU
flush_fp_to_thread(current);
#endif
switch (do_mathemu(regs)) {
case 0:
emulate_single_step(regs);