minix/kernel/arch/i386/do_sdevio.c
Ben Gras 565f13088f make vfs & filesystems use failable copying
Change the kernel to add features to vircopy and safecopies so that
transparent copy fixing won't happen to avoid deadlocks, and such copies
fail with EFAULT.

Transparently making copying work from filesystems (as normally done by
the kernel & VM when copying fails because of missing/readonly memory)
is problematic as it can happen that, for file-mapped ranges, that that
same filesystem that is blocked on the copy request is needed to satisfy
the memory range, leading to deadlock. Dito for VFS itself, if done with
a blocking call.

This change makes the copying done from a filesystem fail in such cases
with EFAULT by VFS adding the CPF_TRY flag to the grants. If a FS call
fails with EFAULT, VFS will then request the range to be made available
to VM after the FS is unblocked, allowing it to be used to satisfy the
range if need be in another VFS thread.

Similarly, for datacopies that VFS itself does, it uses the failable
vircopy variant and callers use a wrapper that talk to VM if necessary
to get the copy to work.

	. kernel: add CPF_TRY flag to safecopies
	. kernel: only request writable ranges to VM for the
	  target buffer when copying fails
	. do copying in VFS TRY-first
	. some fixes in VM to build SANITYCHECK mode
	. add regression test for the cases where
	  - a FS system call needs memory mapped in a process that the
	    FS itself must map.
	  - such a range covers more than one file-mapped region.
	. add 'try' mode to vircopy, physcopy
	. add flags field to copy kernel call messages
	. if CP_FLAG_TRY is set, do not transparently try
	  to fix memory ranges
	. for use by VFS when accessing user buffers to avoid
	  deadlock
	. remove some obsolete backwards compatability assignments
        . VFS: let thread scheduling work for VM requests too
          Allows VFS to make calls to VM while suspending and resuming
          the currently running thread. Does currently not work for the
          main thread.
        . VM: add fix memory range call for use by VFS

Change-Id: I295794269cea51a3163519a9cfe5901301d90b32
2014-07-28 17:05:14 +02:00

163 lines
4.4 KiB
C

/* The kernel call implemented in this file:
* m_type: SYS_SDEVIO
*
* The parameters for this kernel call are:
* m2_i3: DIO_REQUEST (request input or output)
* m2_l1: DIO_PORT (port to read/ write)
* m2_p1: DIO_VEC_ADDR (virtual address of buffer, or grant ID)
* m2_l2: DIO_VEC_SIZE (number of elements)
* m2_i2: DIO_VEC_PROC (process where buffer is)
* m2_i1: DIO_OFFSET (offset into the grant)
*/
#include "kernel/system.h"
#include <minix/devio.h>
#include <minix/endpoint.h>
#include "arch_proto.h"
#if USE_SDEVIO
/*===========================================================================*
* do_sdevio *
*===========================================================================*/
int do_sdevio(struct proc * caller, message *m_ptr)
{
vir_bytes newoffset;
endpoint_t newep;
int proc_nr;
endpoint_t proc_nr_e = m_ptr->DIO_VEC_ENDPT;
vir_bytes count = m_ptr->DIO_VEC_SIZE;
long port = m_ptr->DIO_PORT;
phys_bytes vir_buf;
int i, req_type, req_dir, size, nr_io_range;
struct priv *privp;
struct io_range *iorp;
struct proc *destproc;
int retval;
/* Allow safe copies and accesses to SELF */
if ((m_ptr->DIO_REQUEST & _DIO_SAFEMASK) != _DIO_SAFE &&
proc_nr_e != SELF)
{
static int first= 1;
if (first)
{
first= 0;
printf("do_sdevio: for %d, req %d\n",
m_ptr->m_source, m_ptr->DIO_REQUEST);
}
}
/* Check if process endpoint is OK.
* A driver may directly provide a pointer to a buffer at the user-process
* that initiated the device I/O. Kernel processes, of course, are denied.
*/
if (proc_nr_e == SELF)
okendpt(caller->p_endpoint, &proc_nr);
else
if(!isokendpt(proc_nr_e, &proc_nr))
return(EINVAL);
if (iskerneln(proc_nr)) return(EPERM);
/* Extract direction (in or out) and type (size). */
req_dir = m_ptr->DIO_REQUEST & _DIO_DIRMASK;
req_type = m_ptr->DIO_REQUEST & _DIO_TYPEMASK;
/* Check for 'safe' variants. */
if((m_ptr->DIO_REQUEST & _DIO_SAFEMASK) == _DIO_SAFE) {
/* Map grant address to physical address. */
if(verify_grant(proc_nr_e, caller->p_endpoint,
(cp_grant_id_t) m_ptr->DIO_VEC_ADDR,
count,
req_dir == _DIO_INPUT ? CPF_WRITE : CPF_READ,
(vir_bytes) m_ptr->DIO_OFFSET,
&newoffset, &newep, NULL) != OK) {
printf("do_sdevio: verify_grant failed\n");
return EPERM;
}
if(!isokendpt(newep, &proc_nr))
return(EINVAL);
destproc = proc_addr(proc_nr);
vir_buf = newoffset;
} else {
if(proc_nr != _ENDPOINT_P(caller->p_endpoint))
{
printf("do_sdevio: unsafe sdevio by %d in %d denied\n",
caller->p_endpoint, proc_nr_e);
return EPERM;
}
/* Get and check physical address. */
vir_buf = (phys_bytes) m_ptr->DIO_VEC_ADDR;
destproc = proc_addr(proc_nr);
}
/* current process must be target for phys_* to be OK */
switch_address_space(destproc);
switch (req_type)
{
case _DIO_BYTE: size= 1; break;
case _DIO_WORD: size= 2; break;
case _DIO_LONG: size= 4; break;
default: size= 4; break; /* Be conservative */
}
privp= priv(caller);
if (privp && privp->s_flags & CHECK_IO_PORT)
{
port= m_ptr->DIO_PORT;
nr_io_range= privp->s_nr_io_range;
for (i= 0, iorp= privp->s_io_tab; i<nr_io_range; i++, iorp++)
{
if (port >= iorp->ior_base && port+size-1 <= iorp->ior_limit)
break;
}
if (i >= nr_io_range)
{
printf(
"do_sdevio: I/O port check failed for proc %d, port 0x%x\n",
m_ptr->m_source, port);
retval = EPERM;
goto return_error;
}
}
if (port & (size-1))
{
printf("do_devio: unaligned port 0x%x (size %d)\n", port, size);
retval = EPERM;
goto return_error;
}
/* Perform device I/O for bytes and words. Longs are not supported. */
if (req_dir == _DIO_INPUT) {
switch (req_type) {
case _DIO_BYTE: phys_insb(port, vir_buf, count); break;
case _DIO_WORD: phys_insw(port, vir_buf, count); break;
default:
retval = EINVAL;
goto return_error;
}
} else if (req_dir == _DIO_OUTPUT) {
switch (req_type) {
case _DIO_BYTE: phys_outsb(port, vir_buf, count); break;
case _DIO_WORD: phys_outsw(port, vir_buf, count); break;
default:
retval = EINVAL;
goto return_error;
}
}
else {
retval = EINVAL;
goto return_error;
}
retval = OK;
return_error:
/* switch back to the address of the process which made the call */
switch_address_space(caller);
return retval;
}
#endif /* USE_SDEVIO */