* along with this program; if not, write to the Free Software
* Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
*/
+/*P:450 This file contains the x86-specific lguest code. It used to be all
+ * mixed in with drivers/lguest/core.c but several foolhardy code slashers
+ * wrestled most of the dependencies out to here in preparation for porting
+ * lguest to other architectures (see what I mean by foolhardy?).
+ *
+ * This also contains a couple of non-obvious setup and teardown pieces which
+ * were implemented after days of debugging pain. :*/
#include <linux/kernel.h>
#include <linux/start_kernel.h>
#include <linux/string.h>
(SWITCHER_ADDR + SHARED_SWITCHER_PAGES*PAGE_SIZE))[cpu]);
}
-static DEFINE_PER_CPU(struct lguest *, last_guest);
+static DEFINE_PER_CPU(struct lg_cpu *, last_cpu);
/*S:010
* We approach the Switcher.
*/
static void copy_in_guest_info(struct lg_cpu *cpu, struct lguest_pages *pages)
{
- struct lguest *lg = cpu->lg;
/* Copying all this data can be quite expensive. We usually run the
* same Guest we ran last time (and that Guest hasn't run anywhere else
* meanwhile). If that's not the case, we pretend everything in the
* Guest has changed. */
- if (__get_cpu_var(last_guest) != lg || lg->last_pages != pages) {
- __get_cpu_var(last_guest) = lg;
- lg->last_pages = pages;
- lg->changed = CHANGED_ALL;
+ if (__get_cpu_var(last_cpu) != cpu || cpu->last_pages != pages) {
+ __get_cpu_var(last_cpu) = cpu;
+ cpu->last_pages = pages;
+ cpu->changed = CHANGED_ALL;
}
/* These copies are pretty cheap, so we do them unconditionally: */
/* Set up the two "TSS" members which tell the CPU what stack to use
* for traps which do directly into the Guest (ie. traps at privilege
* level 1). */
- pages->state.guest_tss.esp1 = cpu->esp1;
+ pages->state.guest_tss.sp1 = cpu->esp1;
pages->state.guest_tss.ss1 = cpu->ss1;
/* Copy direct-to-Guest trap entries. */
- if (lg->changed & CHANGED_IDT)
+ if (cpu->changed & CHANGED_IDT)
copy_traps(cpu, pages->state.guest_idt, default_idt_entries);
/* Copy all GDT entries which the Guest can change. */
- if (lg->changed & CHANGED_GDT)
+ if (cpu->changed & CHANGED_GDT)
copy_gdt(cpu, pages->state.guest_gdt);
/* If only the TLS entries have changed, copy them. */
- else if (lg->changed & CHANGED_GDT_TLS)
+ else if (cpu->changed & CHANGED_GDT_TLS)
copy_gdt_tls(cpu, pages->state.guest_gdt);
/* Mark the Guest as unchanged for next time. */
- lg->changed = 0;
+ cpu->changed = 0;
}
/* Finally: the code to actually call into the Switcher to run the Guest. */
{
/* This is a dummy value we need for GCC's sake. */
unsigned int clobber;
- struct lguest *lg = cpu->lg;
/* Copy the guest-specific information into this CPU's "struct
* lguest_pages". */
* 0-th argument above, ie "a"). %ebx contains the
* physical address of the Guest's top-level page
* directory. */
- : "0"(pages), "1"(__pa(lg->pgdirs[cpu->cpu_pgd].pgdir))
+ : "0"(pages), "1"(__pa(cpu->lg->pgdirs[cpu->cpu_pgd].pgdir))
/* We tell gcc that all these registers could change,
* which means we don't have to save and restore them in
* the Switcher. */
* also simplify copy_in_guest_info(). Note that we'd still need to restore
* things when we exit to Launcher userspace, but that's fairly easy.
*
+ * We could also try using this hooks for PGE, but that might be too expensive.
+ *
* The hooks were designed for KVM, but we can also put them to good use. :*/
/*H:040 This is the i386-specific code to setup and run the Guest. Interrupts
* we set it now, so we can trap and pass that trap to the Guest if it
* uses the FPU. */
if (cpu->ts)
- lguest_set_ts();
+ unlazy_fpu(current);
/* SYSENTER is an optimized way of doing system calls. We can't allow
* it because it always jumps to privilege level 0. A normal Guest
* was doing. */
run_guest_once(cpu, lguest_pages(raw_smp_processor_id()));
- /* Note that the "regs" pointer contains two extra entries which are
+ /* Note that the "regs" structure contains two extra entries which are
* not really registers: a trap number which says what interrupt or
* trap made the switcher code come back, and an error code which some
* traps set. */
+ /* Restore SYSENTER if it's supposed to be on. */
+ if (boot_cpu_has(X86_FEATURE_SEP))
+ wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
+
/* If the Guest page faulted, then the cr2 register will tell us the
* bad virtual address. We have to grab this now, because once we
* re-enable interrupts an interrupt could fault and thus overwrite
if (cpu->regs->trapnum == 14)
cpu->arch.last_pagefault = read_cr2();
/* Similarly, if we took a trap because the Guest used the FPU,
- * we have to restore the FPU it expects to see. */
+ * we have to restore the FPU it expects to see.
+ * math_state_restore() may sleep and we may even move off to
+ * a different CPU. So all the critical stuff should be done
+ * before this. */
else if (cpu->regs->trapnum == 7)
math_state_restore();
-
- /* Restore SYSENTER if it's supposed to be on. */
- if (boot_cpu_has(X86_FEATURE_SEP))
- wrmsr(MSR_IA32_SYSENTER_CS, __KERNEL_CS, 0);
}
/*H:130 Now we've examined the hypercall code; our Guest can make requests.
* instructions and skip over it. We return true if we did. */
static int emulate_insn(struct lg_cpu *cpu)
{
- struct lguest *lg = cpu->lg;
u8 insn;
unsigned int insnlen = 0, in = 0, shift = 0;
/* The eip contains the *virtual* address of the Guest's instruction:
return 0;
/* Decoding x86 instructions is icky. */
- insn = lgread(lg, physaddr, u8);
+ insn = lgread(cpu, physaddr, u8);
/* 0x66 is an "operand prefix". It means it's using the upper 16 bits
of the eax register. */
shift = 16;
/* The instruction is 1 byte so far, read the next byte. */
insnlen = 1;
- insn = lgread(lg, physaddr + insnlen, u8);
+ insn = lgread(cpu, physaddr + insnlen, u8);
}
/* We can ignore the lower bit for the moment and decode the 4 opcodes
/*H:050 Once we've re-enabled interrupts, we look at why the Guest exited. */
void lguest_arch_handle_trap(struct lg_cpu *cpu)
{
- struct lguest *lg = cpu->lg;
switch (cpu->regs->trapnum) {
case 13: /* We've intercepted a General Protection Fault. */
/* Check if this was one of those annoying IN or OUT
break;
case 14: /* We've intercepted a Page Fault. */
/* The Guest accessed a virtual address that wasn't mapped.
- * This happens a lot: we don't actually set up most of the
- * page tables for the Guest at all when we start: as it runs
- * it asks for more and more, and we set them up as
- * required. In this case, we don't even tell the Guest that
- * the fault happened.
+ * This happens a lot: we don't actually set up most of the page
+ * tables for the Guest at all when we start: as it runs it asks
+ * for more and more, and we set them up as required. In this
+ * case, we don't even tell the Guest that the fault happened.
*
* The errcode tells whether this was a read or a write, and
* whether kernel or userspace code. */
* Note that if the Guest were really messed up, this could
* happen before it's done the LHCALL_LGUEST_INIT hypercall, so
* lg->lguest_data could be NULL */
- if (lg->lguest_data &&
- put_user(cpu->arch.last_pagefault, &lg->lguest_data->cr2))
- kill_guest(lg, "Writing cr2");
+ if (cpu->lg->lguest_data &&
+ put_user(cpu->arch.last_pagefault,
+ &cpu->lg->lguest_data->cr2))
+ kill_guest(cpu, "Writing cr2");
break;
case 7: /* We've intercepted a Device Not Available fault. */
/* If the Guest doesn't want to know, we already restored the
if (!deliver_trap(cpu, cpu->regs->trapnum))
/* If the Guest doesn't have a handler (either it hasn't
* registered any yet, or it's one of the faults we don't let
- * it handle), it dies with a cryptic error message. */
- kill_guest(lg, "unhandled trap %li at %#lx (%#lx)",
+ * it handle), it dies with this cryptic error message. */
+ kill_guest(cpu, "unhandled trap %li at %#lx (%#lx)",
cpu->regs->trapnum, cpu->regs->eip,
cpu->regs->trapnum == 14 ? cpu->arch.last_pagefault
: cpu->regs->errcode);
* The only exception is the interrupt handlers in switcher.S: their
* addresses are placed in a table (default_idt_entries), so we need to
* update the table with the new addresses. switcher_offset() is a
- * convenience function which returns the distance between the builtin
- * switcher code and the high-mapped copy we just made. */
+ * convenience function which returns the distance between the
+ * compiled-in switcher code and the high-mapped copy we just made. */
for (i = 0; i < IDT_ENTRIES; i++)
default_idt_entries[i] += switcher_offset();
state->guest_gdt_desc.address = (long)&state->guest_gdt;
/* We know where we want the stack to be when the Guest enters
- * the switcher: in pages->regs. The stack grows upwards, so
+ * the Switcher: in pages->regs. The stack grows upwards, so
* we start it at the end of that structure. */
- state->guest_tss.esp0 = (long)(&pages->regs + 1);
+ state->guest_tss.sp0 = (long)(&pages->regs + 1);
/* And this is the GDT entry to use for the stack: we keep a
* couple of special LGUEST entries. */
state->guest_tss.ss0 = LGUEST_DS;
cpu_had_pge = 1;
/* adjust_pge is a helper function which sets or unsets the PGE
* bit on its CPU, depending on the argument (0 == unset). */
- on_each_cpu(adjust_pge, (void *)0, 0, 1);
+ on_each_cpu(adjust_pge, (void *)0, 1);
/* Turn off the feature in the global feature set. */
clear_bit(X86_FEATURE_PGE, boot_cpu_data.x86_capability);
}
if (cpu_had_pge) {
set_bit(X86_FEATURE_PGE, boot_cpu_data.x86_capability);
/* adjust_pge's argument "1" means set PGE. */
- on_each_cpu(adjust_pge, (void *)1, 0, 1);
+ on_each_cpu(adjust_pge, (void *)1, 1);
}
put_online_cpus();
}
int lguest_arch_init_hypercalls(struct lg_cpu *cpu)
{
u32 tsc_speed;
- struct lguest *lg = cpu->lg;
- /* The pointer to the Guest's "struct lguest_data" is the only
- * argument. We check that address now. */
- if (!lguest_address_ok(lg, cpu->hcall->arg1, sizeof(*lg->lguest_data)))
+ /* The pointer to the Guest's "struct lguest_data" is the only argument.
+ * We check that address now. */
+ if (!lguest_address_ok(cpu->lg, cpu->hcall->arg1,
+ sizeof(*cpu->lg->lguest_data)))
return -EFAULT;
/* Having checked it, we simply set lg->lguest_data to point straight
* copy_to_user/from_user from now on, instead of lgread/write. I put
* this in to show that I'm not immune to writing stupid
* optimizations. */
- lg->lguest_data = lg->mem_base + cpu->hcall->arg1;
+ cpu->lg->lguest_data = cpu->lg->mem_base + cpu->hcall->arg1;
/* We insist that the Time Stamp Counter exist and doesn't change with
* cpu frequency. Some devious chip manufacturers decided that TSC
tsc_speed = tsc_khz;
else
tsc_speed = 0;
- if (put_user(tsc_speed, &lg->lguest_data->tsc_khz))
+ if (put_user(tsc_speed, &cpu->lg->lguest_data->tsc_khz))
return -EFAULT;
/* The interrupt code might not like the system call vector. */
- if (!check_syscall_vector(lg))
- kill_guest(lg, "bad syscall vector");
+ if (!check_syscall_vector(cpu->lg))
+ kill_guest(cpu, "bad syscall vector");
return 0;
}
+/*:*/
/*L:030 lguest_arch_setup_regs()
*
git://ftp.safe.ca / safe / jmp / linux-2.6 / blobdiff