d/freebsd-skq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
03efdfd152
freebsd-skq/usr.sbin/ndiscvt/ndiscvt.c

433 lines
11 KiB
C
Raw Normal View History

Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
/*
* Copyright (c) 2003
* Bill Paul <wpaul@windriver.com>. All rights reserved.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by Bill Paul.
* 4. Neither the name of the author nor the names of any co-contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY Bill Paul AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL Bill Paul OR THE VOICES IN HIS HEAD
* BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
* CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
* SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
* INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
* CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
* ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF
* THE POSSIBILITY OF SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/types.h>
#include <sys/queue.h>
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
#include <sys/socket.h>
#include <net/if.h>
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
#include <stdlib.h>
Clean up and apply the fix for PR 83477. The calculation for locating the start of the section headers has to take into account the fact that the image_nt_header is really variable sized. It happens that the existing calculation is correct for _most_ production binaries produced by the Windows DDK, but if we get a binary with oddball offsets, the PE loader could crash. Changes from the supplied patch are: - We don't really need to use the IMAGE_SIZEOF_NT_HEADER() macro when computing how much of the header to return to callers of pe_get_optional_header(). While it's important to take the variable size of the header into account in other calculations, we never actually look at anything outside the non-variable portion of the header. This saves callers from having to allocate a variable sized buffer off the heap (I purposely tried to avoid using malloc() in subr_pe.c to make it easier to compile in both the -D_KERNEL and !-D_KERNEL case), and since we're copying into a buffer on the stack, we always have to copy the same amount of data or else we'll trash the stack something fierce. - We need <stddef.h> to get offsetof() in the !-D_KERNEL case. - ndiscvt.c needs the IMAGE_FIRST_SECTION() macro too, since it does a little bit of section pre-processing. PR: kern/83477
2005-10-26 18:46:27 +00:00
#include <stddef.h>
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
#include <unistd.h>
#include <stdio.h>
#include <errno.h>
#include <string.h>
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
#include <libgen.h>
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
#include <err.h>
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
#include <ctype.h>
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
#include <compat/ndis/pe_var.h>
#include "inf.h"
static int insert_padding(void **, int *);
extern const char *__progname;
/*
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
* Sections within Windows PE files are defined using virtual
* and physical address offsets and virtual and physical sizes.
* The physical values define how the section data is stored in
* the executable file while the virtual values describe how the
* sections will look once loaded into memory. It happens that
* the linker in the Microsoft(r) DDK will tend to generate
* binaries where the virtual and physical values are identical,
* which means in most cases we can just transfer the file
* directly to memory without any fixups. This is not always
* the case though, so we have to be prepared to handle files
* where the in-memory section layout differs from the disk file
* section layout.
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
*
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
* There are two kinds of variations that can occur: the relative
* virtual address of the section might be different from the
* physical file offset, and the virtual section size might be
* different from the physical size (for example, the physical
* size of the .data section might be 1024 bytes, but the virtual
* size might be 1384 bytes, indicating that the data section should
* actually use up 1384 bytes in RAM and be padded with zeros). What we
* do is read the original file into memory and then make an in-memory
* copy with all of the sections relocated, re-sized and zero padded
* according to the virtual values specified in the section headers.
* We then emit the fixed up image file for use by the if_ndis driver.
* This way, we don't have to do the fixups inside the kernel.
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
*/
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
#define ROUND_DOWN(n, align) (((uintptr_t)n) & ~((align) - 1l))
#define ROUND_UP(n, align) ROUND_DOWN(((uintptr_t)n) + (align) - 1l, \
(align))
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
#define SET_HDRS(x) \
dos_hdr = (image_dos_header *)x; \
nt_hdr = (image_nt_header *)(x + dos_hdr->idh_lfanew); \
Clean up and apply the fix for PR 83477. The calculation for locating the start of the section headers has to take into account the fact that the image_nt_header is really variable sized. It happens that the existing calculation is correct for _most_ production binaries produced by the Windows DDK, but if we get a binary with oddball offsets, the PE loader could crash. Changes from the supplied patch are: - We don't really need to use the IMAGE_SIZEOF_NT_HEADER() macro when computing how much of the header to return to callers of pe_get_optional_header(). While it's important to take the variable size of the header into account in other calculations, we never actually look at anything outside the non-variable portion of the header. This saves callers from having to allocate a variable sized buffer off the heap (I purposely tried to avoid using malloc() in subr_pe.c to make it easier to compile in both the -D_KERNEL and !-D_KERNEL case), and since we're copying into a buffer on the stack, we always have to copy the same amount of data or else we'll trash the stack something fierce. - We need <stddef.h> to get offsetof() in the !-D_KERNEL case. - ndiscvt.c needs the IMAGE_FIRST_SECTION() macro too, since it does a little bit of section pre-processing. PR: kern/83477
2005-10-26 18:46:27 +00:00
sect_hdr = IMAGE_FIRST_SECTION(nt_hdr);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
static
int insert_padding(imgbase, imglen)
void **imgbase;
int *imglen;
{
image_section_header *sect_hdr;
image_dos_header *dos_hdr;
image_nt_header *nt_hdr;
image_optional_header opt_hdr;
int i = 0, sections, curlen = 0;
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
int offaccum = 0, oldraddr, oldrlen;
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
uint8_t *newimg, *tmp;
newimg = malloc(*imglen);
if (newimg == NULL)
return(ENOMEM);
bcopy(*imgbase, newimg, *imglen);
curlen = *imglen;
if (pe_get_optional_header((vm_offset_t)newimg, &opt_hdr))
return(0);
sections = pe_numsections((vm_offset_t)newimg);
SET_HDRS(newimg);
for (i = 0; i < sections; i++) {
oldraddr = sect_hdr->ish_rawdataaddr;
oldrlen = sect_hdr->ish_rawdatasize;
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
sect_hdr->ish_rawdataaddr = sect_hdr->ish_vaddr;
offaccum += ROUND_UP(sect_hdr->ish_vaddr - oldraddr,
opt_hdr.ioh_filealign);
offaccum +=
ROUND_UP(sect_hdr->ish_misc.ish_vsize,
opt_hdr.ioh_filealign) -
ROUND_UP(sect_hdr->ish_rawdatasize,
opt_hdr.ioh_filealign);
tmp = realloc(newimg, *imglen + offaccum);
if (tmp == NULL) {
free(newimg);
return(ENOMEM);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
}
Today, RealTek sent me a driver to test which had been compiled with some debug support turned on. It turns out the sections in this driver binary had relative virtual addresses (RVAs) that were different from the raw addresses, and it had a .data section where the virtual size was much larger than the raw size. (Most production binaries produced with the Microsoft DDK have RVA == PA.) There's code in the ndiscvt(8) utility that's supposed to handle the vsize != rsize case, but it turns out it was slightly broken, and it failed to handle the RVA != RA case at all. Hopefully, this commit will fix all that.
2004-08-02 18:54:01 +00:00
newimg = tmp;
SET_HDRS(newimg);
sect_hdr += i;
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
bzero(newimg + sect_hdr->ish_rawdataaddr,
ROUND_UP(sect_hdr->ish_misc.ish_vsize,
opt_hdr.ioh_filealign));
bcopy((uint8_t *)(*imgbase) + oldraddr,
newimg + sect_hdr->ish_rawdataaddr, oldrlen);
sect_hdr++;
}
free(*imgbase);
*imgbase = newimg;
*imglen += offaccum;
return(0);
}
static void
usage(void)
{
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
fprintf(stderr, "Usage: %s [-O] [-i <inffile>] -s <sysfile> "
Fix usage() (-d is really -n).
2004-01-02 23:35:57 +00:00
"[-n devname] [-o outfile]\n", __progname);
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
fprintf(stderr, " %s -f <firmfile>\n", __progname);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
exit(1);
}
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
static void
bincvt(char *sysfile, char *outfile, void *img, int fsize)
{
char *ptr;
char tname[] = "/tmp/ndiscvt.XXXXXX";
char sysbuf[1024];
FILE *binfp;
mkstemp(tname);
binfp = fopen(tname, "a+");
if (binfp == NULL)
err(1, "opening %s failed", tname);
if (fwrite(img, fsize, 1, binfp) != 1)
err(1, "failed to output binary image");
fclose(binfp);
outfile = strdup(basename(outfile));
if (strchr(outfile, '.'))
*strchr(outfile, '.') = '\0';
snprintf(sysbuf, sizeof(sysbuf),
Fix a small bug in firmcvt: outfile must be strdup()ed. Also, add conditional code to allow different invokations for objcopy depending on whether we're compiled on an i386 arch or amd64 arch, so that we can produce x86-64 object files on amd64.
2005-02-19 07:37:01 +00:00
#ifdef __i386__
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
"objcopy -I binary -O elf32-i386-freebsd -B i386 %s %s.o\n",
Fix a small bug in firmcvt: outfile must be strdup()ed. Also, add conditional code to allow different invokations for objcopy depending on whether we're compiled on an i386 arch or amd64 arch, so that we can produce x86-64 object files on amd64.
2005-02-19 07:37:01 +00:00
#endif
#ifdef __amd64__
"objcopy -I binary -O elf64-x86-64 -B i386 %s %s.o\n",
#endif
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
tname, outfile);
printf("%s", sysbuf);
system(sysbuf);
unlink(tname);
ptr = tname;
while (*ptr) {
if (*ptr == '/' || *ptr == '.')
*ptr = '_';
ptr++;
}
snprintf(sysbuf, sizeof(sysbuf),
Use a ndis_ prefix on the C variable instead of directly using the .sys filename, this would fail if the filename started with a number. PR: bin/84911 Submitted by: Fredrik Lindberg
2008-04-15 04:17:13 +00:00
"objcopy --redefine-sym _binary_%s_start=ndis_%s_drv_data_start "
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
"--strip-symbol _binary_%s_size "
Use a ndis_ prefix on the C variable instead of directly using the .sys filename, this would fail if the filename started with a number. PR: bin/84911 Submitted by: Fredrik Lindberg
2008-04-15 04:17:13 +00:00
"--redefine-sym _binary_%s_end=ndis_%s_drv_data_end %s.o %s.o\n",
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
tname, sysfile, tname, tname, sysfile, outfile, outfile);
printf("%s", sysbuf);
system(sysbuf);
return;
}
static void
firmcvt(char *firmfile)
{
char *basefile, *outfile, *ptr;
char sysbuf[1024];
Fix a small bug in firmcvt: outfile must be strdup()ed. Also, add conditional code to allow different invokations for objcopy depending on whether we're compiled on an i386 arch or amd64 arch, so that we can produce x86-64 object files on amd64.
2005-02-19 07:37:01 +00:00
outfile = strdup(basename(firmfile));
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
basefile = strdup(outfile);
snprintf(sysbuf, sizeof(sysbuf),
Fix a small bug in firmcvt: outfile must be strdup()ed. Also, add conditional code to allow different invokations for objcopy depending on whether we're compiled on an i386 arch or amd64 arch, so that we can produce x86-64 object files on amd64.
2005-02-19 07:37:01 +00:00
#ifdef __i386__
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
"objcopy -I binary -O elf32-i386-freebsd -B i386 %s %s.o\n",
Fix a small bug in firmcvt: outfile must be strdup()ed. Also, add conditional code to allow different invokations for objcopy depending on whether we're compiled on an i386 arch or amd64 arch, so that we can produce x86-64 object files on amd64.
2005-02-19 07:37:01 +00:00
#endif
#ifdef __amd64__
"objcopy -I binary -O elf64-x86-64 -B i386 %s %s.o\n",
#endif
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
firmfile, outfile);
printf("%s", sysbuf);
system(sysbuf);
ptr = firmfile;
while (*ptr) {
if (*ptr == '/' || *ptr == '.')
*ptr = '_';
ptr++;
}
ptr = basefile;
while (*ptr) {
if (*ptr == '/' || *ptr == '.')
*ptr = '_';
else
*ptr = tolower(*ptr);
ptr++;
}
snprintf(sysbuf, sizeof(sysbuf),
"objcopy --redefine-sym _binary_%s_start=%s_start "
"--strip-symbol _binary_%s_size "
"--redefine-sym _binary_%s_end=%s_end %s.o %s.o\n",
firmfile, basefile, firmfile, firmfile,
basefile, outfile, outfile);
ptr = sysbuf;
printf("%s", sysbuf);
system(sysbuf);
snprintf(sysbuf, sizeof(sysbuf),
"ld -Bshareable -d -warn-common -o %s.ko %s.o\n",
outfile, outfile);
printf("%s", sysbuf);
system(sysbuf);
free(basefile);
exit(0);
}
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
int
main(int argc, char *argv[])
{
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
FILE *fp, *outfp;
int i, bin = 0;
void *img;
int n, fsize, cnt;
unsigned char *ptr;
char *inffile = NULL, *sysfile = NULL;
char *outfile = NULL, *firmfile = NULL;
char *dname = NULL;
int ch;
while((ch = getopt(argc, argv, "i:s:o:n:f:O")) != -1) {
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
switch(ch) {
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
case 'f':
firmfile = optarg;
break;
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
case 'i':
inffile = optarg;
break;
case 's':
sysfile = optarg;
break;
case 'o':
outfile = optarg;
break;
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
case 'n':
dname = optarg;
break;
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
case 'O':
bin = 1;
break;
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
default:
usage();
break;
}
}
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
if (firmfile != NULL)
firmcvt(firmfile);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
if (sysfile == NULL)
usage();
/* Open the .SYS file and load it into memory */
fp = fopen(sysfile, "r");
if (fp == NULL)
err(1, "opening .SYS file '%s' failed", sysfile);
fseek (fp, 0L, SEEK_END);
fsize = ftell (fp);
rewind (fp);
img = calloc(fsize, 1);
n = fread (img, fsize, 1, fp);
fclose(fp);
if (insert_padding(&img, &fsize)) {
fprintf(stderr, "section relocation failed\n");
exit(1);
}
if (outfile == NULL || strcmp(outfile, "-") == 0)
outfp = stdout;
else {
outfp = fopen(outfile, "w");
if (outfp == NULL)
err(1, "opening output file '%s' failed", outfile);
}
fprintf(outfp, "\n/*\n");
fprintf(outfp, " * Generated from %s and %s (%d bytes)\n",
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
inffile == NULL ? "<notused>" : inffile, sysfile, fsize);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
fprintf(outfp, " */\n\n");
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
if (dname != NULL) {
if (strlen(dname) > IFNAMSIZ)
err(1, "selected device name '%s' is "
"too long (max chars: %d)", dname, IFNAMSIZ);
fprintf (outfp, "#define NDIS_DEVNAME \"%s\"\n", dname);
Handle WinNT .inf files with a $windows nt$ signature but no .NT decorated AddReg sections. Also insert extra newline after emitting device name overrides.
2004-01-02 21:13:21 +00:00
fprintf (outfp, "#define NDIS_MODNAME %s\n\n", dname);
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
}
if (inffile == NULL) {
Add preliminary support for PCMCIA devices in addition to PCI/cardbus. if_ndis.c has been split into if_ndis_pci.c and if_ndis_pccard.c. The ndiscvt(8) utility should be able to parse device info for PCMCIA devices now. The ndis_alloc_amem() has moved from kern_ndis.c to if_ndis_pccard.c so that kern_ndis.c no longer depends on pccard. NOTE: this stuff is not guaranteed to work 100% correctly yet. So far I have been able to load/init my PCMCIA Cisco Aironet 340 card, but it crashes in the interrupt handler. The existing support for PCI/cardbus devices should still work as before.
2004-03-07 02:49:06 +00:00
fprintf (outfp, "#ifdef NDIS_REGVALS\n");
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
fprintf (outfp, "ndis_cfg ndis_regvals[] = {\n");
Clean up ndiscvt a bit (leaving out the -i flag didn't work) and add copyrights to the inf parser files. Add a -n flag to ndiscvt to allow the user to override the default device name of NDIS devices. Instead of "ndis0, ndis1, etc..." you can have "foo0, foo1, etc..." This allows you to have more than one kind of NDIS device in the kernel at the same time. Convert from printf() to device_printf() in if_ndis.c, kern_ndis.c and subr_ndis.c. Create UMA zones for ndis_packet and ndis_buffer structs allocated on transmit. The zones are created and destroyed in the modevent handler in kern_ndis.c. printf() and UMA changes submitted by green@freebsd.org
2004-01-02 04:31:06 +00:00
fprintf (outfp, "\t{ NULL, NULL, { 0 }, 0 }\n");
Add preliminary support for PCMCIA devices in addition to PCI/cardbus. if_ndis.c has been split into if_ndis_pci.c and if_ndis_pccard.c. The ndiscvt(8) utility should be able to parse device info for PCMCIA devices now. The ndis_alloc_amem() has moved from kern_ndis.c to if_ndis_pccard.c so that kern_ndis.c no longer depends on pccard. NOTE: this stuff is not guaranteed to work 100% correctly yet. So far I have been able to load/init my PCMCIA Cisco Aironet 340 card, but it crashes in the interrupt handler. The existing support for PCI/cardbus devices should still work as before.
2004-03-07 02:49:06 +00:00
fprintf (outfp, "#endif /* NDIS_REGVALS */\n");
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
fprintf (outfp, "};\n\n");
} else {
fp = fopen(inffile, "r");
if (fp == NULL)
err(1, "opening .INF file '%s' failed", inffile);
inf_parse(fp, outfp);
fclose(fp);
}
Add preliminary support for PCMCIA devices in addition to PCI/cardbus. if_ndis.c has been split into if_ndis_pci.c and if_ndis_pccard.c. The ndiscvt(8) utility should be able to parse device info for PCMCIA devices now. The ndis_alloc_amem() has moved from kern_ndis.c to if_ndis_pccard.c so that kern_ndis.c no longer depends on pccard. NOTE: this stuff is not guaranteed to work 100% correctly yet. So far I have been able to load/init my PCMCIA Cisco Aironet 340 card, but it crashes in the interrupt handler. The existing support for PCI/cardbus devices should still work as before.
2004-03-07 02:49:06 +00:00
fprintf(outfp, "\n#ifdef NDIS_IMAGE\n");
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
if (bin) {
sysfile = strdup(basename(sysfile));
Silence GCC4 signed/unsigned pointer mismatch warning.
2006-09-21 01:48:47 +00:00
ptr = (unsigned char *)sysfile;
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
while (*ptr) {
if (*ptr == '.')
*ptr = '_';
ptr++;
}
fprintf(outfp,
Use a ndis_ prefix on the C variable instead of directly using the .sys filename, this would fail if the filename started with a number. PR: bin/84911 Submitted by: Fredrik Lindberg
2008-04-15 04:17:13 +00:00
"\nextern unsigned char ndis_%s_drv_data_start[];\n",
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
sysfile);
fprintf(outfp, "static unsigned char *drv_data = "
Use a ndis_ prefix on the C variable instead of directly using the .sys filename, this would fail if the filename started with a number. PR: bin/84911 Submitted by: Fredrik Lindberg
2008-04-15 04:17:13 +00:00
"ndis_%s_drv_data_start;\n\n", sysfile);
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
bincvt(sysfile, outfile, img, fsize);
goto done;
}
Add preliminary support for PCMCIA devices in addition to PCI/cardbus. if_ndis.c has been split into if_ndis_pci.c and if_ndis_pccard.c. The ndiscvt(8) utility should be able to parse device info for PCMCIA devices now. The ndis_alloc_amem() has moved from kern_ndis.c to if_ndis_pccard.c so that kern_ndis.c no longer depends on pccard. NOTE: this stuff is not guaranteed to work 100% correctly yet. So far I have been able to load/init my PCMCIA Cisco Aironet 340 card, but it crashes in the interrupt handler. The existing support for PCI/cardbus devices should still work as before.
2004-03-07 02:49:06 +00:00
fprintf(outfp, "\nextern unsigned char drv_data[];\n\n");
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
fprintf(outfp, "__asm__(\".data\");\n");
Next step on the road to IRPs: create and use an imitation of the Windows DRIVER_OBJECT and DEVICE_OBJECT mechanism so that we can simulate driver stacking. In Windows, each loaded driver image is attached to a DRIVER_OBJECT structure. Windows uses the registry to match up a given vendor/device ID combination with a corresponding DRIVER_OBJECT. When a driver image is first loaded, its DriverEntry() routine is invoked, which sets up the AddDevice() function pointer in the DRIVER_OBJECT and creates a dispatch table (based on IRP major codes). When a Windows bus driver detects a new device, it creates a Physical Device Object (PDO) for it. This is a DEVICE_OBJECT structure, with semantics analagous to that of a device_t in FreeBSD. The Windows PNP manager will invoke the driver's AddDevice() function and pass it pointers to the DRIVER_OBJECT and the PDO. The AddDevice() function then creates a new DRIVER_OBJECT structure of its own. This is known as the Functional Device Object (FDO) and corresponds roughly to a private softc instance. The driver uses IoAttachDeviceToDeviceStack() to add this device object to the driver stack for this PDO. Subsequent drivers (called filter drivers in Windows-speak) can be loaded which add themselves to the stack. When someone issues an IRP to a device, it travel along the stack passing through several possible filter drivers until it reaches the functional driver (which actually knows how to talk to the hardware) at which point it will be completed. This is how Windows achieves driver layering. Project Evil now simulates most of this. if_ndis now has a modevent handler which will use MOD_LOAD and MOD_UNLOAD events to drive the creation and destruction of DRIVER_OBJECTs. (The load event also does the relocation/dynalinking of the image.) We don't have a registry, so the DRIVER_OBJECTS are stored in a linked list for now. Eventually, the list entry will contain the vendor/device ID list extracted from the .INF file. When ndis_probe() is called and detectes a supported device, it will create a PDO for the device instance and attach it to the DRIVER_OBJECT just as in Windows. ndis_attach() will then call our NdisAddDevice() handler to create the FDO. The NDIS miniport block is now a device extension hung off the FDO, just as it is in Windows. The miniport characteristics table is now an extension hung off the DRIVER_OBJECT as well (the characteristics are the same for all devices handled by a given driver, so they don't need to be per-instance.) We also do an IoAttachDeviceToDeviceStack() to put the FDO on the stack for the PDO. There are a couple of fake bus drivers created for the PCI and pccard buses. Eventually, there will be one for USB, which will actually accept USB IRP.s Things should still work just as before, only now we do things in the proper order and maintain the correct framework to support passing IRPs between drivers. Various changes: - corrected the comments about IRQL handling in subr_hal.c to more accurately reflect reality - update ndiscvt to make the drv_data symbol in ndis_driver_data.h a global so that if_ndis_pci.o and/or if_ndis_pccard.o can see it. - Obtain the softc pointer from the miniport block by referencing the PDO rather than a private pointer of our own (nmb_ifp is no longer used) - implement IoAttachDeviceToDeviceStack(), IoDetachDevice(), IoGetAttachedDevice(), IoAllocateDriverObjectExtension(), IoGetDriverObjectExtension(), IoCreateDevice(), IoDeleteDevice(), IoAllocateIrp(), IoReuseIrp(), IoMakeAssociatedIrp(), IoFreeIrp(), IoInitializeIrp() - fix a few mistakes in the driver_object and device_object definitions - add a new module, kern_windrv.c, to handle the driver registration and relocation/dynalinkign duties (which don't really belong in kern_ndis.c). - made ndis_block and ndis_chars in the ndis_softc stucture pointers and modified all references to it - fixed NdisMRegisterMiniport() and NdisInitializeWrapper() so they work correctly with the new driver_object mechanism - changed ndis_attach() to call NdisAddDevice() instead of ndis_load_driver() (which is now deprecated) - used ExAllocatePoolWithTag()/ExFreePool() in lookaside list routines instead of kludged up alloc/free routines - added kern_windrv.c to sys/modules/ndis/Makefile and files.i386.
2005-02-08 17:23:25 +00:00
fprintf(outfp, "__asm__(\".globl drv_data\");\n");
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
fprintf(outfp, "__asm__(\".type drv_data, @object\");\n");
fprintf(outfp, "__asm__(\".size drv_data, %d\");\n", fsize);
fprintf(outfp, "__asm__(\"drv_data:\");\n");
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
ptr = img;
cnt = 0;
while(cnt < fsize) {
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
fprintf (outfp, "__asm__(\".byte ");
for (i = 0; i < 10; i++) {
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
cnt++;
if (cnt == fsize) {
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
fprintf(outfp, "0x%.2X\");\n", ptr[i]);
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
goto done;
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
} else {
if (i == 9)
fprintf(outfp, "0x%.2X\");\n", ptr[i]);
else
fprintf(outfp, "0x%.2X, ", ptr[i]);
}
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
}
Make ndiscvt(8) emit the binary image array as inline assembly code rather than a char array. Emitting the data as a big char array works fine in the typical case, where a .sys file may be ~50K in size. Unfortunately, some .sys files can be several hundred Kbytes in size, or even several megabytes in size. One extreme case is the Intel centrino wireless driver, which is 2.4MB. This causes us to emit an ndis_driver_data.h file that's on the order of 15MB in size, and gcc consumes enormous amounts of virtual memory while trying to compile it. On my laptop, with 128MB of RAM and 256MB of swap space, gcc consumed all available VM and crashed without being able to compile if_ndis.o. By emitting the array as assembler, we bypass the C compiler and consume much less memory. I was able to easily test compile if_ndis.ko with the centrino driver on my laptop after this change. This is merely a convenience, and should not have any operational effect on the NDISulator itself.
2003-12-18 21:47:14 +00:00
ptr += 10;
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
}
done:
Add preliminary support for PCMCIA devices in addition to PCI/cardbus. if_ndis.c has been split into if_ndis_pci.c and if_ndis_pccard.c. The ndiscvt(8) utility should be able to parse device info for PCMCIA devices now. The ndis_alloc_amem() has moved from kern_ndis.c to if_ndis_pccard.c so that kern_ndis.c no longer depends on pccard. NOTE: this stuff is not guaranteed to work 100% correctly yet. So far I have been able to load/init my PCMCIA Cisco Aironet 340 card, but it crashes in the interrupt handler. The existing support for PCI/cardbus devices should still work as before.
2004-03-07 02:49:06 +00:00
fprintf(outfp, "#endif /* NDIS_IMAGE */\n");
Big mess 'o changes: - Give ndiscvt(8) the ability to process a .SYS file directly into a .o file so that we don't have to emit big messy char arrays into the ndis_driver_data.h file. This behavior is currently optional, but may become the default some day. - Give ndiscvt(8) the ability to turn arbitrary files into .ko files so that they can be pre-loaded or kldloaded. (Both this and the previous change involve using objcopy(1)). - Give NdisOpenFile() the ability to 'read' files out of kernel memory that have been kldloaded or pre-loaded, and disallow the use of the normal vn_open() file opening method during bootstrap (when no filesystems have been mounted yet). Some people have reported that kldloading if_ndis.ko works fine when the system is running multiuser but causes a panic when the modile is pre-loaded by /boot/loader. This happens with drivers that need to use NdisOpenFile() to access external files (i.e. firmware images). NdisOpenFile() won't work during kernel bootstrapping because no filesystems have been mounted. To get around this, you can now do the following: o Say you have a firmware file called firmware.img o Do: ndiscvt -f firmware.img -- this creates firmware.img.ko o Put the firmware.img.ko in /boot/kernel o add firmware.img_load="YES" in /boot/loader.conf o add if_ndis_load="YES" and ndis_load="YES" as well Now the loader will suck the additional file into memory as a .ko. The phony .ko has two symbols in it: filename_start and filename_end, which are generated by objcopy(1). ndis_open_file() will traverse each module in the module list looking for these symbols and, if it finds them, it'll use them to generate the file mapping address and length values that the caller of NdisOpenFile() wants. As a bonus, this will even work if the file has been statically linked into the kernel itself, since the "kernel" module is searched too. (ndiscvt(8) will generate both filename.o and filename.ko for you). - Modify the mechanism used to provide make-pretend FASTCALL support. Rather than using inline assembly to yank the first two arguments out of %ecx and %edx, we now use the __regparm__(3) attribute (and the __stdcall__ attribute) and use some macro magic to re-order the arguments and provide dummy arguments as needed so that the arguments passed in registers end up in the right place. Change taken from DragonflyBSD version of the NDISulator.
2004-08-01 20:04:31 +00:00
Commit the ndiscvt(8) utility too. (Missed it in the last import.)
2003-12-11 22:38:14 +00:00
if (fp != NULL)
fclose(fp);
fclose(outfp);
free(img);
exit(0);
}
Reference in New Issue Copy Permalink