genode/repos/base-hw/src/core/include/spec/x86/cpu_support.h

259 lines
6.9 KiB
C
Raw Normal View History

/*
* \brief x86 CPU driver for core
* \author Adrian-Ken Rueegsegger
* \author Martin stein
* \author Reto Buerki
* \date 2015-02-06
*/
/*
* Copyright (C) 2015 Genode Labs GmbH
*
* This file is part of the Genode OS framework, which is distributed
* under the terms of the GNU General Public License version 2.
*/
#ifndef _SPEC__X86__CPU_SUPPORT_H_
#define _SPEC__X86__CPU_SUPPORT_H_
2015-03-20 12:22:34 +01:00
/* Genode includes */
#include <util/register.h>
#include <kernel/interface_support.h>
#include <cpu/cpu_state.h>
2015-03-20 12:22:34 +01:00
/* base includes */
#include <unmanaged_singleton.h>
/* core includes */
#include <fpu.h>
#include <gdt.h>
#include <idt.h>
#include <tss.h>
#include <timer.h>
extern int _mt_idt;
extern int _mt_tss;
namespace Genode { class Cpu; }
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
class Genode::Cpu
{
public:
2015-03-20 12:22:34 +01:00
struct Pd {};
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
struct Cr0; /* Control register 0 */
struct Cr2; /* Control register 2 */
struct Cr3; /* Control register 3 */
struct Cr4; /* Control register 4 */
/**
* Extend basic CPU state by members relevant for 'base-hw' only
*/
struct Context : Cpu_state
{
hw_x86_64: Implementation of IA-32e paging IA-32e paging translates 48-bit linear addresses to 52-bit physical addresses. Translation structures are hierarchical and four levels deep. The current implementation supports regular 4KB and 1 GB and 2 MB large page mappings. Memory typing is not yet implemented since the encoded type bits depend on the active page attribute table (PAT)*. For detailed information refer to Intel SDM Vol. 3A, section 4.5. * The default PAT after power up does not allow the encoding of the write-combining memory type, see Intel SDM Vol. 3A, section 11.12.4. * Add common IA-32e paging descriptor type: The type represents a table entry and encompasses all fields shared by paging structure entries of all four levels (PML4, PDPT, PD and PT). * Simplify PT entry type by using common descriptor: Differing fields are the physical address, the global flag and the memory type flags. * Simplify directory entry type by using common descriptor: Page directory entries (PDPT and PD) have an additional 'page size' field that specifies if the entry references a next level paging structure or represents a large page mapping. * Simplify PML4 entry type by using common descriptor Top-level paging structure entries (PML4) do not have a 'pat' flag and the memory type is specified by the 'pwt' and 'pcd' fields only. * Implement access right merging for directory paging entries The access rights for translations are determined by the U/S, R/W and XD flags. Paging structure entries that reference other tables must provide the superset of rights required for all entries of the referenced table. Thus merge access rights of new mappings into existing directory entries to grant additional rights if needed. * Add cr3 register definition: The control register 3 is used to set the current page-directory base register. * Add cr3 variable to x86_64 Cpu Context The variable designates the address of the top-level paging structure. * Return current cr3 value as translation table base * Set context cr3 value on translation table assignment * Implement switch to virtual mode in kernel Activate translation table in init_virt_kernel function by updating the cr3 register. * Ignore accessed and dirty flags when comparing existing table entries These flags can be set by the MMU and must be disregarded.
2015-02-16 15:20:06 +01:00
/**
* Address of top-level paging structure.
*/
addr_t cr3;
/**
* Return base of assigned translation table
*/
hw_x86_64: Implementation of IA-32e paging IA-32e paging translates 48-bit linear addresses to 52-bit physical addresses. Translation structures are hierarchical and four levels deep. The current implementation supports regular 4KB and 1 GB and 2 MB large page mappings. Memory typing is not yet implemented since the encoded type bits depend on the active page attribute table (PAT)*. For detailed information refer to Intel SDM Vol. 3A, section 4.5. * The default PAT after power up does not allow the encoding of the write-combining memory type, see Intel SDM Vol. 3A, section 11.12.4. * Add common IA-32e paging descriptor type: The type represents a table entry and encompasses all fields shared by paging structure entries of all four levels (PML4, PDPT, PD and PT). * Simplify PT entry type by using common descriptor: Differing fields are the physical address, the global flag and the memory type flags. * Simplify directory entry type by using common descriptor: Page directory entries (PDPT and PD) have an additional 'page size' field that specifies if the entry references a next level paging structure or represents a large page mapping. * Simplify PML4 entry type by using common descriptor Top-level paging structure entries (PML4) do not have a 'pat' flag and the memory type is specified by the 'pwt' and 'pcd' fields only. * Implement access right merging for directory paging entries The access rights for translations are determined by the U/S, R/W and XD flags. Paging structure entries that reference other tables must provide the superset of rights required for all entries of the referenced table. Thus merge access rights of new mappings into existing directory entries to grant additional rights if needed. * Add cr3 register definition: The control register 3 is used to set the current page-directory base register. * Add cr3 variable to x86_64 Cpu Context The variable designates the address of the top-level paging structure. * Return current cr3 value as translation table base * Set context cr3 value on translation table assignment * Implement switch to virtual mode in kernel Activate translation table in init_virt_kernel function by updating the cr3 register. * Ignore accessed and dirty flags when comparing existing table entries These flags can be set by the MMU and must be disregarded.
2015-02-16 15:20:06 +01:00
addr_t translation_table() const { return cr3; }
/**
* Initialize context
*
* \param table physical base of appropriate translation table
* \param core whether it is a core thread or not
*/
void init(addr_t const table, bool core);
};
/**
* An usermode execution state
*/
struct User_context : Context, Fpu::Context
{
/**
* Support for kernel calls
*/
void user_arg_0(Kernel::Call_arg const arg) { rdi = arg; }
void user_arg_1(Kernel::Call_arg const arg) { rsi = arg; }
void user_arg_2(Kernel::Call_arg const arg) { rdx = arg; }
void user_arg_3(Kernel::Call_arg const arg) { rcx = arg; }
void user_arg_4(Kernel::Call_arg const arg) { r8 = arg; }
void user_arg_5(Kernel::Call_arg const arg) { r9 = arg; }
void user_arg_6(Kernel::Call_arg const arg) { r10 = arg; }
void user_arg_7(Kernel::Call_arg const arg) { r11 = arg; }
Kernel::Call_arg user_arg_0() const { return rdi; }
Kernel::Call_arg user_arg_1() const { return rsi; }
Kernel::Call_arg user_arg_2() const { return rdx; }
Kernel::Call_arg user_arg_3() const { return rcx; }
Kernel::Call_arg user_arg_4() const { return r8; }
Kernel::Call_arg user_arg_5() const { return r9; }
Kernel::Call_arg user_arg_6() const { return r10; }
Kernel::Call_arg user_arg_7() const { return r11; }
};
protected:
Fpu _fpu;
Idt * _idt;
Tss * _tss;
public:
Cpu();
Fpu & fpu() { return _fpu; }
static constexpr addr_t exception_entry = 0xffff0000;
static constexpr addr_t mtc_size = 1 << 13;
/**
* Wait for the next interrupt as cheap as possible
*/
static void wait_for_interrupt() { asm volatile ("pause"); }
/**
* Return wether to retry an undefined user instruction after this call
*/
bool retry_undefined_instr(Context&) { return false; }
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
/**
* Return kernel name of the executing CPU
*/
static unsigned executing_id() { return 0; }
/**
* Return kernel name of the primary CPU
*/
static unsigned primary_id() { return 0; }
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
/**
* Switch to new context
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
*
* \param context next CPU context
hw_x86_64: Enable FPU support * Enable the use of the FXSAVE and FXRSTOR instructions, see Intel SDM Vol. 3C, section 2.5. * The state of the x87 floating point unit (FPU) is loaded and saved on demand. * Make the cr0 control register accessible in the Cpu class. This is in preparation of the upcoming FPU management. * Access to the FPU is disabled by setting the Task Switch flag in the cr0 register. * Access to the FPU is enabled by clearing the Task Switch flag in the cr0 register. * Implement FPU initialization * Add is_fpu_enabled helper function * Add pointer to CPU lazy state to CPU class * Init FPU when finishing kernel initialization * Add function to retry FPU instruction: Similar to the ARM mechanism to retry undefined instructions, implement a function for retrying an FPU instruction. If a floating-point instruction causes an #NM exception due to the FPU being disabled, it can be retried after the correct FPU state is restored, saving the current state and enabling the FPU in the process. * Disable FPU when switching to different user context: This enables lazy save/restore of the FPU since trying to execute a floating point instruction when the FPU is disabled will cause a #NM exception. * Declare constant for #NM exception * Retry FPU instruction on #NM exception * Assure alignment of FXSAVE area: The FXSAVE area is 512-byte memory region that must be 16-byte aligned. As it turns out the alignment attribute is not honored in all cases so add a workaround to assure the alignment constraint is met by manually rounding the start of the FXSAVE area to the next 16-byte boundary if necessary.
2015-03-06 14:09:35 +01:00
*/
void switch_to(User_context &context) { _fpu.switch_to(context); }
};
struct Genode::Cpu::Cr0 : Register<64>
{
struct Pe : Bitfield< 0, 1> { }; /* Protection Enable */
struct Mp : Bitfield< 1, 1> { }; /* Monitor Coprocessor */
struct Em : Bitfield< 2, 1> { }; /* Emulation */
struct Ts : Bitfield< 3, 1> { }; /* Task Switched */
struct Et : Bitfield< 4, 1> { }; /* Extension Type */
struct Ne : Bitfield< 5, 1> { }; /* Numeric Error */
struct Wp : Bitfield<16, 1> { }; /* Write Protect */
struct Am : Bitfield<18, 1> { }; /* Alignment Mask */
struct Nw : Bitfield<29, 1> { }; /* Not Write-through */
struct Cd : Bitfield<30, 1> { }; /* Cache Disable */
struct Pg : Bitfield<31, 1> { }; /* Paging */
static void write(access_t const v) {
asm volatile ("mov %0, %%cr0" :: "r" (v) : ); }
static access_t read()
{
access_t v;
asm volatile ("mov %%cr0, %0" : "=r" (v) :: );
return v;
}
};
/**
* Control register 2: Page-fault linear address
*
* See Intel SDM Vol. 3A, section 2.5.
*/
struct Genode::Cpu::Cr2 : Register<64>
{
struct Addr : Bitfield<0, 63> { };
static access_t read()
{
access_t v;
asm volatile ("mov %%cr2, %0" : "=r" (v) :: );
return v;
}
};
/**
* Control register 3: Page-Directory base register
*
* See Intel SDM Vol. 3A, section 2.5.
*/
struct Genode::Cpu::Cr3 : Register<64>
{
struct Pwt : Bitfield<3,1> { }; /* Page-level write-through */
struct Pcd : Bitfield<4,1> { }; /* Page-level cache disable */
struct Pdb : Bitfield<12, 36> { }; /* Page-directory base address */
static void write(access_t const v) {
asm volatile ("mov %0, %%cr3" :: "r" (v) : ); }
static access_t read()
{
access_t v;
asm volatile ("mov %%cr3, %0" : "=r" (v) :: );
return v;
}
/**
* Return initialized value
*
* \param table base of targeted translation table
*/
static access_t init(addr_t const table) {
return Pdb::masked(table); }
};
struct Genode::Cpu::Cr4 : Register<64>
{
struct Vme : Bitfield< 0, 1> { }; /* Virtual-8086 Mode Extensions */
struct Pvi : Bitfield< 1, 1> { }; /* Protected-Mode Virtual IRQs */
struct Tsd : Bitfield< 2, 1> { }; /* Time Stamp Disable */
struct De : Bitfield< 3, 1> { }; /* Debugging Exceptions */
struct Pse : Bitfield< 4, 1> { }; /* Page Size Extensions */
struct Pae : Bitfield< 5, 1> { }; /* Physical Address Extension */
struct Mce : Bitfield< 6, 1> { }; /* Machine-Check Enable */
struct Pge : Bitfield< 7, 1> { }; /* Page Global Enable */
struct Pce : Bitfield< 8, 1> { }; /* Performance-Monitoring Counter
Enable*/
struct Osfxsr : Bitfield< 9, 1> { }; /* OS Support for FXSAVE and
FXRSTOR instructions*/
struct Osxmmexcpt : Bitfield<10, 1> { }; /* OS Support for Unmasked
SIMD/FPU Exceptions */
struct Vmxe : Bitfield<13, 1> { }; /* VMX Enable */
struct Smxe : Bitfield<14, 1> { }; /* SMX Enable */
struct Fsgsbase : Bitfield<16, 1> { }; /* FSGSBASE-Enable */
struct Pcide : Bitfield<17, 1> { }; /* PCIDE Enable */
struct Osxsave : Bitfield<18, 1> { }; /* XSAVE and Processor Extended
States-Enable */
struct Smep : Bitfield<20, 1> { }; /* SMEP Enable */
struct Smap : Bitfield<21, 1> { }; /* SMAP Enable */
static void write(access_t const v) {
asm volatile ("mov %0, %%cr4" :: "r" (v) : ); }
static access_t read()
{
access_t v;
asm volatile ("mov %%cr4, %0" : "=r" (v) :: );
return v;
}
};
#endif /* _SPEC__X86__CPU_SUPPORT_H_ */