system/freebsd-src
system/freebsd-src
1
0
Fork 0
mirror of https://github.com/freebsd/freebsd-src synced 2024-07-22 10:48:02 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity
freebsd-src/stand/common/load_elf_obj.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

581 lines
15 KiB
C
Raw Normal View History

Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
/*-
* Copyright (c) 2004 Ian Dowse <iedowse@freebsd.org>
* Copyright (c) 1998 Michael Smith <msmith@freebsd.org>
* Copyright (c) 1998 Peter Wemm <peter@freebsd.org>
* 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR 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 THE AUTHOR OR CONTRIBUTORS 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/param.h>
#include <sys/exec.h>
#include <sys/linker.h>
#include <sys/module.h>
Prefer stdint.h to inttypes.h since the added prototypes form the latter aren't used. Prefer sys/link_elf.h to link.h so we're only dependent on the kernel tree. The default installation of link.h just includes this file, and any benefit from that is outweighed by the hassle it causes. This reduces the footprint of files needed from the system includes (or sysroot in buildworld). Sponsored by: Netflix
2017-12-05 21:38:04 +00:00
#include <stdint.h>
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
#include <string.h>
#include <machine/elf.h>
#include <stand.h>
Prefer stdint.h to inttypes.h since the added prototypes form the latter aren't used. Prefer sys/link_elf.h to link.h so we're only dependent on the kernel tree. The default installation of link.h just includes this file, and any benefit from that is outweighed by the hassle it causes. This reduces the footprint of files needed from the system includes (or sysroot in buildworld). Sponsored by: Netflix
2017-12-05 21:38:04 +00:00
#include <sys/link_elf.h>
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
#include "bootstrap.h"
#define COPYOUT(s,d,l) archsw.arch_copyout((vm_offset_t)(s), d, l)
#if defined(__i386__) && __ELF_WORD_SIZE == 64
#undef ELF_TARG_CLASS
#undef ELF_TARG_MACH
#define ELF_TARG_CLASS ELFCLASS64
#define ELF_TARG_MACH EM_X86_64
#endif
typedef struct elf_file {
Elf_Ehdr hdr;
Elf_Shdr *e_shdr;
int symtabindex; /* Index of symbol table */
int shstrindex; /* Index of section name string table */
int fd;
vm_offset_t off;
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
#ifdef LOADER_VERIEXEC_VECTX
struct vectx *vctx;
#endif
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
} *elf_file_t;
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
#ifdef LOADER_VERIEXEC_VECTX
#define VECTX_HANDLE(ef) (ef)->vctx
#else
#define VECTX_HANDLE(ef) (ef)->fd
#endif
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
static int __elfN(obj_loadimage)(struct preloaded_file *mp, elf_file_t ef,
Prefer uintXX_t to u_intXX_t A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. -- Ralph Waldo Emerson
2018-03-13 16:33:00 +00:00
uint64_t loadaddr);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
static int __elfN(obj_lookup_set)(struct preloaded_file *mp, elf_file_t ef,
const char *name, Elf_Addr *startp, Elf_Addr *stopp, int *countp);
static int __elfN(obj_reloc_ptr)(struct preloaded_file *mp, elf_file_t ef,
Elf_Addr p, void *val, size_t len);
static int __elfN(obj_parse_modmetadata)(struct preloaded_file *mp,
elf_file_t ef);
Make our ELF64 type definitions match standards. In particular this means: o Remove Elf64_Quarter, o Redefine Elf64_Half to be 16-bit, o Redefine Elf64_Word to be 32-bit, o Add Elf64_Xword and Elf64_Sxword for 64-bit entities, o Use Elf_Size in MI code to abstract the difference between Elf32_Word and Elf64_Word. o Add Elf_Ssize as the signed counterpart of Elf_Size. MFC after: 2 weeks
2005-12-18 04:52:37 +00:00
static Elf_Addr __elfN(obj_symaddr)(struct elf_file *ef, Elf_Size symidx);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
const char *__elfN(obj_kerneltype) = "elf kernel";
const char *__elfN(obj_moduletype) = "elf obj module";
/*
* Attempt to load the file (file) as an ELF module. It will be stored at
* (dest), and a pointer to a module structure describing the loaded object
* will be saved in (result).
*/
int
Prefer uintXX_t to u_intXX_t A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. -- Ralph Waldo Emerson
2018-03-13 16:33:00 +00:00
__elfN(obj_loadfile)(char *filename, uint64_t dest,
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
struct preloaded_file **result)
{
struct preloaded_file *fp, *kfp;
struct elf_file ef;
Elf_Ehdr *hdr;
int err;
ssize_t bytes_read;
fp = NULL;
bzero(&ef, sizeof(struct elf_file));
/*
* Open the image, read and validate the ELF header
*/
if (filename == NULL) /* can't handle nameless */
return(EFTYPE);
if ((ef.fd = open(filename, O_RDONLY)) == -1)
return(errno);
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
#ifdef LOADER_VERIEXEC_VECTX
{
int verror;
ef.vctx = vectx_open(ef.fd, filename, 0L, NULL, &verror, __func__);
if (verror) {
printf("Unverified %s: %s\n", filename, ve_error_get());
close(ef.fd);
free(ef.vctx);
return (EAUTH);
}
}
#endif
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
hdr = &ef.hdr;
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
bytes_read = VECTX_READ(VECTX_HANDLE(&ef), hdr, sizeof(*hdr));
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
if (bytes_read != sizeof(*hdr)) {
err = EFTYPE; /* could be EIO, but may be small file */
goto oerr;
}
/* Is it ELF? */
if (!IS_ELF(*hdr)) {
err = EFTYPE;
goto oerr;
}
if (hdr->e_ident[EI_CLASS] != ELF_TARG_CLASS || /* Layout ? */
hdr->e_ident[EI_DATA] != ELF_TARG_DATA ||
hdr->e_ident[EI_VERSION] != EV_CURRENT || /* Version ? */
hdr->e_version != EV_CURRENT ||
hdr->e_machine != ELF_TARG_MACH || /* Machine ? */
hdr->e_type != ET_REL) {
err = EFTYPE;
goto oerr;
}
if (hdr->e_shnum * hdr->e_shentsize == 0 || hdr->e_shoff == 0 ||
hdr->e_shentsize != sizeof(Elf_Shdr)) {
err = EFTYPE;
goto oerr;
}
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
#if defined(LOADER_VERIEXEC) && !defined(LOADER_VERIEXEC_VECTX)
if (verify_file(ef.fd, filename, bytes_read, VE_MUST, __func__) < 0) {
err = EAUTH;
goto oerr;
Enable veriexec for loader This relies on libbearssl and libsecureboot to verify files read by loader in a maner equivalent to how mac_veriexec Note: disabled by default. Use is initially expected to be by embeded vendors Reviewed by: emaste, imp Sponsored by: Juniper Networks Differential Revision: D16336
2019-02-26 06:22:10 +00:00
}
#endif
loader: implement multiboot support for Xen Dom0 Implement a subset of the multiboot specification in order to boot Xen and a FreeBSD Dom0 from the FreeBSD bootloader. This multiboot implementation is tailored to boot Xen and FreeBSD Dom0, and it will most surely fail to boot any other multiboot compilant kernel. In order to detect and boot the Xen microkernel, two new file formats are added to the bootloader, multiboot and multiboot_obj. Multiboot support must be tested before regular ELF support, since Xen is a multiboot kernel that also uses ELF. After a multiboot kernel is detected, all the other loaded kernels/modules are parsed by the multiboot_obj format. The layout of the loaded objects in memory is the following; first the Xen kernel is loaded as a 32bit ELF into memory (Xen will switch to long mode by itself), after that the FreeBSD kernel is loaded as a RAW file (Xen will parse and load it using it's internal ELF loader), and finally the metadata and the modules are loaded using the native FreeBSD way. After everything is loaded we jump into Xen's entry point using a small trampoline. The order of the multiboot modules passed to Xen is the following, the first module is the RAW FreeBSD kernel, and the second module is the metadata and the FreeBSD modules. Since Xen will relocate the memory position of the second multiboot module (the one that contains the metadata and native FreeBSD modules), we need to stash the original modulep address inside of the metadata itself in order to recalculate its position once booted. This also means the metadata must come before the loaded modules, so after loading the FreeBSD kernel a portion of memory is reserved in order to place the metadata before booting. In order to tell the loader to boot Xen and then the FreeBSD kernel the following has to be added to the /boot/loader.conf file: xen_cmdline="dom0_mem=1024M dom0_max_vcpus=2 dom0pvh=1 console=com1,vga" xen_kernel="/boot/xen" The first argument contains the command line that will be passed to the Xen kernel, while the second argument is the path to the Xen kernel itself. This can also be done manually from the loader command line, by for example typing the following set of commands: OK unload OK load /boot/xen dom0_mem=1024M dom0_max_vcpus=2 dom0pvh=1 console=com1,vga OK load kernel OK load zfs OK load if_tap OK load ... OK boot Sponsored by: Citrix Systems R&D Reviewed by: jhb Differential Revision: https://reviews.freebsd.org/D517 For the Forth bits: Submitted by: Julien Grall <julien.grall AT citrix.com>
2015-01-15 16:27:20 +00:00
kfp = file_findfile(NULL, __elfN(obj_kerneltype));
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
if (kfp == NULL) {
printf("elf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadfile: can't load module before kernel\n");
err = EPERM;
goto oerr;
}
Add 2 new archsw interfaces: 1. arch_loadaddr - used by platform code to adjust the address at which the object gets loaded. Implement PC98 using this new interface instead of using conditional compilation. For ELF objects the ELF header is passed as the data pointer. For raw files it's the filename. Note that ELF objects are first considered as raw files. 2. arch_loadseg - used by platform code to keep track of actual segments, so that (instruction) caches can be flushed or translations can be created. Both the ELF header as well as the program header are passed to allow platform code to treat the kernel proper differently from any additional modules and to have all the relevant details of the loaded segment (e.g. protection).
2011-04-03 22:31:51 +00:00
if (archsw.arch_loadaddr != NULL)
dest = archsw.arch_loadaddr(LOAD_ELF, hdr, dest);
else
dest = roundup(dest, PAGE_SIZE);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
/*
* Ok, we think we should handle this.
*/
fp = file_alloc();
if (fp == NULL) {
printf("elf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadfile: cannot allocate module info\n");
err = EPERM;
goto out;
}
fp->f_name = strdup(filename);
fp->f_type = strdup(__elfN(obj_moduletype));
Add support for module_verbose Set module_verbose to control the printing of information about loaded modules and kernel: 0 MODULE_VERBOSE_SILENT None 1 MODULE_VERBOSE_SIZE Pathname and size 2 MODULE_VERBOSE_TWIDDLE as for 1 but also twiddle for progress 3 MODULE_VERBOSE_FULL extra detail When the loader is verifying modules we already have a running indication of progress and module_verbose=0 makes sense. Reviewed by: rpokala Differential Revision: https://reviews.freebsd.org/D34245
2022-02-13 20:45:57 +00:00
if (module_verbose > MODULE_VERBOSE_SILENT)
printf("%s ", filename);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
fp->f_size = __elfN(obj_loadimage)(fp, &ef, dest);
if (fp->f_size == 0 || fp->f_addr == 0)
goto ioerr;
/* save exec header as metadata */
file_addmetadata(fp, MODINFOMD_ELFHDR, sizeof(*hdr), hdr);
/* Load OK, return module pointer */
*result = (struct preloaded_file *)fp;
err = 0;
goto out;
ioerr:
err = EIO;
oerr:
file_discard(fp);
out:
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
#ifdef LOADER_VERIEXEC_VECTX
if (!err && ef.vctx) {
int verror;
verror = vectx_close(ef.vctx, VE_MUST, __func__);
if (verror) {
err = EAUTH;
file_discard(fp);
}
}
#endif
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
close(ef.fd);
if (ef.e_shdr != NULL)
free(ef.e_shdr);
return(err);
}
/*
* With the file (fd) open on the image, and (ehdr) containing
* the Elf header, load the image at (off)
*/
static int
Prefer uintXX_t to u_intXX_t A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. -- Ralph Waldo Emerson
2018-03-13 16:33:00 +00:00
__elfN(obj_loadimage)(struct preloaded_file *fp, elf_file_t ef, uint64_t off)
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
{
Elf_Ehdr *hdr;
Minimize backward seeks when trying to load ELF relocatable modules. Some of loader filesystems are very ill equipped to handle seeking backwards within the file. Namely, tftp requires trasfer to be restarted from the start of the file every time we go backwards.
2011-06-19 13:35:41 +00:00
Elf_Shdr *shdr, *cshdr, *lshdr;
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
vm_offset_t firstaddr, lastaddr;
int i, nsym, res, ret, shdrbytes, symstrindex;
ret = 0;
firstaddr = lastaddr = (vm_offset_t)off;
hdr = &ef->hdr;
ef->off = (vm_offset_t)off;
/* Read in the section headers. */
shdrbytes = hdr->e_shnum * hdr->e_shentsize;
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
shdr = alloc_pread(VECTX_HANDLE(ef), (off_t)hdr->e_shoff, shdrbytes);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
if (shdr == NULL) {
printf("\nelf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadimage: read section headers failed\n");
goto out;
}
ef->e_shdr = shdr;
/*
* Decide where to load everything, but don't read it yet.
* We store the load address as a non-zero sh_addr value.
* Start with the code/data and bss.
*/
for (i = 0; i < hdr->e_shnum; i++)
shdr[i].sh_addr = 0;
for (i = 0; i < hdr->e_shnum; i++) {
completely ignore zero-sized elf sections in modules of elf object type (amd64) Current code doesn't check size of elf sections and may perform needless actions of zero-sized memory allocation and similar. The bigger issue is that alignment requirement of a zero-sized section gets effectively applied to the next section if it has smaller alignment requirement. But other tools, like gdb and consequently kgdb, completely ignore zero-sized sections and thus may map symbols to addresses differently. Zero-sized sections are not typical in general. Their typical (only, even) cause in FreeBSD modules is inline assembly that creates custom sections which is found in pcpu.h and vnet.h. Mere inclusion of one of those header files produces a custom section in elf output. If there is no actual use for the section in a given module, then the section remains empty. Better solution is to avoid creating zero-sized sections altogether, which is in plans. Preloaded modules are handled in boot code (load_elf_obj.c), while dynamically loaded modules are handled by kernel (link_elf_obj.c). Based on code by: np MFC after: 3 weeks
2010-07-23 17:07:51 +00:00
if (shdr[i].sh_size == 0)
continue;
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
switch (shdr[i].sh_type) {
case SHT_PROGBITS:
case SHT_NOBITS:
Since kernel modules can now contain sections of type SHT_AMD64_UNWIND, the boot loader should not skip over these anymore while loading images. Otherwise the kernel can still panic when it doesn't find the .eh_frame section belonging to the .rela.eh_frame section. Unfortunately this will require installing boot loaders from sys/boot before attempting to boot with a new kernel. Reviewed by: kib MFC after: 2 weeks X-MFC-With: r296419
2016-03-06 15:57:43 +00:00
#if defined(__i386__) || defined(__amd64__)
Use the ABI-prescribed name for SHT_X86_64_UNWIND in the loader and kernel linker, after the r297686. Sponsored by: The FreeBSD Foundation
2016-04-08 10:23:48 +00:00
case SHT_X86_64_UNWIND:
Since kernel modules can now contain sections of type SHT_AMD64_UNWIND, the boot loader should not skip over these anymore while loading images. Otherwise the kernel can still panic when it doesn't find the .eh_frame section belonging to the .rela.eh_frame section. Unfortunately this will require installing boot loaders from sys/boot before attempting to boot with a new kernel. Reviewed by: kib MFC after: 2 weeks X-MFC-With: r296419
2016-03-06 15:57:43 +00:00
#endif
stand: Load INIT_ARRAY and FINI_ARRAY sections This is required for preloading modules into a KASAN-configured kernel. MFC after: 1 week Sponsored by: The FreeBSD Foundation
2021-03-16 19:01:41 +00:00
case SHT_INIT_ARRAY:
case SHT_FINI_ARRAY:
Require the SHF_ALLOC flag for program sections from kernel object modules. ELF object files can contain program sections which are not supposed to be loaded into memory (e.g. .comment). Normally the static linker uses these flags to decide which sections are allocated to loadable program segments in ELF binaries and shared objects (including kernels on all architectures and kernel modules on architectures other than amd64). Mapping ELF object files (such as amd64 kernel modules) into memory directly is a bit of a grey area. ELF object files are intended to be used as inputs to the static linker. As a result, there is not a standardized definition for what the memory layout of an ELF object should be (none of the section headers have valid virtual memory addresses for example). The kernel and loader were not checking the SHF_ALLOC flag but loading any program sections with certain types such as SHT_PROGBITS. As a result, the kernel and loader would load into RAM some sections that weren't marked with SHF_ALLOC such as .comment that are not loaded into RAM for kernel modules on other architectures (which are implemented as ELF shared objects). Aside from possibly requiring slightly more RAM to hold a kernel module this does not affect runtime correctness as the kernel relocates symbols based on the layout it uses. Debuggers such as gdb and lldb do not extract symbol tables from a running process or kernel. Instead, they replicate the memory layout of ELF executables and shared objects and use that to construct their own symbol tables. For executables and shared objects this works fine. For ELF objects the current logic in kgdb (and probably lldb based on a simple reading) assumes that only sections with SHF_ALLOC are memory resident when constructing a memory layout. If the debugger constructs a different memory layout than the kernel, then it will compute different addresses for symbols causing symbols in the debugger to appear to have the wrong values (though the kernel itself is working fine). The current port of mdb does not check SHF_ALLOC as it replicates the kernel's logic in its existing kernel support. The bfd linker sorts the sections in ELF object files such that all of the allocated sections (sections with SHF_ALLOCATED) are placed first followed by unallocated sections. As a result, when kgdb composed a memory layout using only the allocated sections, this layout happened to match the layout used by the kernel and loader. The lld linker does not sort the sections in ELF object files and mixed allocated and unallocated sections. This resulted in kgdb composing a different memory layout than the kernel and loader. We could either patch kgdb (and possibly in the future lldb) to use custom handling when generating memory layouts for kernel modules that are ELF objects, or we could change the kernel and loader to check SHF_ALLOCATED. I chose the latter as I feel we shouldn't be loading things into RAM that the module won't use. This should mostly be a NOP when linking with bfd but will allow the existing kgdb to work with amd64 kernel modules linked with lld. Note that we only require SHF_ALLOC for "program" sections for types like SHT_PROGBITS and SHT_NOBITS. Other section types such as symbol tables, string tables, and relocations must also be loaded and are not marked with SHF_ALLOC. Reported by: np Reviewed by: kib, emaste MFC after: 1 month Sponsored by: Chelsio Communications Differential Revision: https://reviews.freebsd.org/D13926
2018-01-17 22:51:59 +00:00
if ((shdr[i].sh_flags & SHF_ALLOC) == 0)
break;
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
lastaddr = roundup(lastaddr, shdr[i].sh_addralign);
shdr[i].sh_addr = (Elf_Addr)lastaddr;
lastaddr += shdr[i].sh_size;
break;
}
}
/* Symbols. */
nsym = 0;
for (i = 0; i < hdr->e_shnum; i++) {
switch (shdr[i].sh_type) {
case SHT_SYMTAB:
nsym++;
ef->symtabindex = i;
break;
}
}
if (nsym != 1) {
printf("\nelf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadimage: file has no valid symbol table\n");
goto out;
}
lastaddr = roundup(lastaddr, shdr[ef->symtabindex].sh_addralign);
shdr[ef->symtabindex].sh_addr = (Elf_Addr)lastaddr;
lastaddr += shdr[ef->symtabindex].sh_size;
symstrindex = shdr[ef->symtabindex].sh_link;
if (symstrindex < 0 || symstrindex >= hdr->e_shnum ||
shdr[symstrindex].sh_type != SHT_STRTAB) {
printf("\nelf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadimage: file has invalid symbol strings\n");
goto out;
}
lastaddr = roundup(lastaddr, shdr[symstrindex].sh_addralign);
shdr[symstrindex].sh_addr = (Elf_Addr)lastaddr;
lastaddr += shdr[symstrindex].sh_size;
/* Section names. */
if (hdr->e_shstrndx == 0 || hdr->e_shstrndx >= hdr->e_shnum ||
shdr[hdr->e_shstrndx].sh_type != SHT_STRTAB) {
printf("\nelf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadimage: file has no section names\n");
goto out;
}
ef->shstrindex = hdr->e_shstrndx;
lastaddr = roundup(lastaddr, shdr[ef->shstrindex].sh_addralign);
shdr[ef->shstrindex].sh_addr = (Elf_Addr)lastaddr;
lastaddr += shdr[ef->shstrindex].sh_size;
/* Relocation tables. */
for (i = 0; i < hdr->e_shnum; i++) {
switch (shdr[i].sh_type) {
case SHT_REL:
case SHT_RELA:
Ignore relocation tables for non-memory-resident sections. As a followup to r328101, ignore relocation tables for ELF object sections that are not memory resident. For modules loaded by the loader, ignore relocation tables whose associated section was not loaded by the loader (sh_addr is zero). For modules loaded at runtime via kldload(2), ignore relocation tables whose associated section is not marked with SHF_ALLOC. Reported by: Mori Hiroki <yamori813@yahoo.co.jp>, adrian Tested on: mips, mips64 MFC after: 1 month Sponsored by: DARPA / AFRL
2018-02-05 23:35:33 +00:00
if ((shdr[shdr[i].sh_info].sh_flags & SHF_ALLOC) == 0)
break;
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
lastaddr = roundup(lastaddr, shdr[i].sh_addralign);
shdr[i].sh_addr = (Elf_Addr)lastaddr;
lastaddr += shdr[i].sh_size;
break;
}
}
/* Clear the whole area, including bss regions. */
kern_bzero(firstaddr, lastaddr - firstaddr);
Minimize backward seeks when trying to load ELF relocatable modules. Some of loader filesystems are very ill equipped to handle seeking backwards within the file. Namely, tftp requires trasfer to be restarted from the start of the file every time we go backwards.
2011-06-19 13:35:41 +00:00
/* Figure section with the lowest file offset we haven't loaded yet. */
for (cshdr = NULL; /* none */; /* none */)
{
/*
* Find next section to load. The complexity of this loop is
* O(n^2), but with the number of sections being typically
* small, we do not care.
*/
lshdr = cshdr;
for (i = 0; i < hdr->e_shnum; i++) {
if (shdr[i].sh_addr == 0 ||
shdr[i].sh_type == SHT_NOBITS)
continue;
/* Skip sections that were loaded already. */
if (lshdr != NULL &&
lshdr->sh_offset >= shdr[i].sh_offset)
continue;
/* Find section with smallest offset. */
if (cshdr == lshdr ||
cshdr->sh_offset > shdr[i].sh_offset)
cshdr = &shdr[i];
}
if (cshdr == lshdr)
break;
veloader use vectx API for kernel and modules The vectx API, computes the hash for verifying a file as it is read. This avoids the overhead of reading files twice - once to verify, then again to load. For doing an install via loader, avoiding the need to rewind large files is critical. This API is only used for modules, kernel and mdimage as these are the biggest files read by the loader. The reduction in boot time depends on how expensive the I/O is on any given platform. On a fast VM we see 6% improvement. For install via loader the first file to be verified is likely to be the kernel, so some of the prep work (finding manifest etc) done by verify_file() needs to be factored so it can be reused for vectx_open(). For missing or unrecognized fingerprint entries, we fail in vectx_open() unless verifying is disabled. Otherwise fingerprint check happens in vectx_close() and since this API is only used for files which must be verified (VE_MUST) we panic if we get an incorrect hash. Reviewed by: imp,tsoome MFC after: 1 week Sponsored by: Juniper Networks Differential Revision: https://reviews.freebsd.org//D23827
2020-03-08 17:42:42 +00:00
if (kern_pread(VECTX_HANDLE(ef), (vm_offset_t)cshdr->sh_addr,
Minimize backward seeks when trying to load ELF relocatable modules. Some of loader filesystems are very ill equipped to handle seeking backwards within the file. Namely, tftp requires trasfer to be restarted from the start of the file every time we go backwards.
2011-06-19 13:35:41 +00:00
cshdr->sh_size, (off_t)cshdr->sh_offset) != 0) {
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
printf("\nelf" __XSTRING(__ELF_WORD_SIZE)
"_obj_loadimage: read failed\n");
goto out;
}
}
file_addmetadata(fp, MODINFOMD_SHDR, shdrbytes, shdr);
res = __elfN(obj_parse_modmetadata)(fp, ef);
if (res != 0)
goto out;
ret = lastaddr - firstaddr;
fp->f_addr = firstaddr;
Add support for module_verbose Set module_verbose to control the printing of information about loaded modules and kernel: 0 MODULE_VERBOSE_SILENT None 1 MODULE_VERBOSE_SIZE Pathname and size 2 MODULE_VERBOSE_TWIDDLE as for 1 but also twiddle for progress 3 MODULE_VERBOSE_FULL extra detail When the loader is verifying modules we already have a running indication of progress and module_verbose=0 makes sense. Reviewed by: rpokala Differential Revision: https://reviews.freebsd.org/D34245
2022-02-13 20:45:57 +00:00
if (module_verbose > MODULE_VERBOSE_SILENT)
printf("size 0x%lx at 0x%lx", (u_long)ret, (u_long)firstaddr);
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
out:
Add support for module_verbose Set module_verbose to control the printing of information about loaded modules and kernel: 0 MODULE_VERBOSE_SILENT None 1 MODULE_VERBOSE_SIZE Pathname and size 2 MODULE_VERBOSE_TWIDDLE as for 1 but also twiddle for progress 3 MODULE_VERBOSE_FULL extra detail When the loader is verifying modules we already have a running indication of progress and module_verbose=0 makes sense. Reviewed by: rpokala Differential Revision: https://reviews.freebsd.org/D34245
2022-02-13 20:45:57 +00:00
if (module_verbose > MODULE_VERBOSE_SILENT)
printf("\n");
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
return ret;
}
#if defined(__i386__) && __ELF_WORD_SIZE == 64
struct mod_metadata64 {
int md_version; /* structure version MDTV_* */
int md_type; /* type of entry MDT_* */
Prefer uintXX_t to u_intXX_t A foolish consistency is the hobgoblin of little minds, adored by little statesmen and philosophers and divines. With consistency a great soul has simply nothing to do. -- Ralph Waldo Emerson
2018-03-13 16:33:00 +00:00
uint64_t md_data; /* specific data */
uint64_t md_cval; /* common string label */
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
};
#endif
int
__elfN(obj_parse_modmetadata)(struct preloaded_file *fp, elf_file_t ef)
{
struct mod_metadata md;
#if defined(__i386__) && __ELF_WORD_SIZE == 64
struct mod_metadata64 md64;
#endif
struct mod_depend *mdepend;
struct mod_version mver;
char *s;
int error, modcnt, minfolen;
Elf_Addr v, p, p_stop;
if (__elfN(obj_lookup_set)(fp, ef, "modmetadata_set", &p, &p_stop,
&modcnt) != 0)
Don't return an error if a kld does not contain any modules (e.g. a kld that only contained a sysctl). The kernel linker allows such modules, so the boot loader should not reject them. MFC after: 2 weeks
2012-06-20 21:06:51 +00:00
return 0;
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
modcnt = 0;
while (p < p_stop) {
COPYOUT(p, &v, sizeof(v));
error = __elfN(obj_reloc_ptr)(fp, ef, p, &v, sizeof(v));
if (error != 0)
return (error);
#if defined(__i386__) && __ELF_WORD_SIZE == 64
COPYOUT(v, &md64, sizeof(md64));
error = __elfN(obj_reloc_ptr)(fp, ef, v, &md64, sizeof(md64));
if (error != 0)
return (error);
md.md_version = md64.md_version;
md.md_type = md64.md_type;
md.md_cval = (const char *)(uintptr_t)md64.md_cval;
md.md_data = (void *)(uintptr_t)md64.md_data;
#else
COPYOUT(v, &md, sizeof(md));
error = __elfN(obj_reloc_ptr)(fp, ef, v, &md, sizeof(md));
if (error != 0)
return (error);
#endif
p += sizeof(Elf_Addr);
switch(md.md_type) {
case MDT_DEPEND:
s = strdupout((vm_offset_t)md.md_cval);
minfolen = sizeof(*mdepend) + strlen(s) + 1;
mdepend = malloc(minfolen);
if (mdepend == NULL)
return ENOMEM;
COPYOUT((vm_offset_t)md.md_data, mdepend,
sizeof(*mdepend));
strcpy((char*)(mdepend + 1), s);
free(s);
file_addmetadata(fp, MODINFOMD_DEPLIST, minfolen,
mdepend);
free(mdepend);
break;
case MDT_VERSION:
s = strdupout((vm_offset_t)md.md_cval);
COPYOUT((vm_offset_t)md.md_data, &mver, sizeof(mver));
file_addmodule(fp, s, mver.mv_version, NULL);
free(s);
modcnt++;
break;
case MDT_MODULE:
Reserve and ignore the a new module metadata type MDT_PNP_INFO for associating an optional PNP hint table with this module. In the future, when these are added, these changes will silently ignore the new type they would otherwise warn about. It will always be safe to ignore this data. Get this into the builds today for some future proofing. MFC After: 3 days
2015-01-15 00:46:30 +00:00
case MDT_PNP_INFO:
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
break;
default:
printf("unknown type %d\n", md.md_type);
break;
}
}
return 0;
}
static int
__elfN(obj_lookup_set)(struct preloaded_file *fp, elf_file_t ef,
const char* name, Elf_Addr *startp, Elf_Addr *stopp, int *countp)
{
Elf_Ehdr *hdr;
Elf_Shdr *shdr;
char *p;
vm_offset_t shstrtab;
int i;
hdr = &ef->hdr;
shdr = ef->e_shdr;
shstrtab = shdr[ef->shstrindex].sh_addr;
for (i = 0; i < hdr->e_shnum; i++) {
if (shdr[i].sh_type != SHT_PROGBITS)
continue;
if (shdr[i].sh_name == 0)
continue;
p = strdupout(shstrtab + shdr[i].sh_name);
if (strncmp(p, "set_", 4) == 0 && strcmp(p + 4, name) == 0) {
*startp = shdr[i].sh_addr;
*stopp = shdr[i].sh_addr + shdr[i].sh_size;
*countp = (*stopp - *startp) / sizeof(Elf_Addr);
free(p);
return (0);
}
free(p);
}
return (ESRCH);
}
/*
* Apply any intra-module relocations to the value. p is the load address
* of the value and val/len is the value to be modified. This does NOT modify
* the image in-place, because this is done by kern_linker later on.
*/
static int
__elfN(obj_reloc_ptr)(struct preloaded_file *mp, elf_file_t ef, Elf_Addr p,
void *val, size_t len)
{
Elf_Ehdr *hdr;
Elf_Shdr *shdr;
Elf_Addr off = p;
Elf_Addr base;
Elf_Rela a, *abase;
Elf_Rel r, *rbase;
int error, i, j, nrel, nrela;
hdr = &ef->hdr;
shdr = ef->e_shdr;
for (i = 0; i < hdr->e_shnum; i++) {
if (shdr[i].sh_type != SHT_RELA && shdr[i].sh_type != SHT_REL)
continue;
base = shdr[shdr[i].sh_info].sh_addr;
if (base == 0 || shdr[i].sh_addr == 0)
continue;
if (off < base || off + len > base +
shdr[shdr[i].sh_info].sh_size)
continue;
switch (shdr[i].sh_type) {
case SHT_RELA:
abase = (Elf_Rela *)(intptr_t)shdr[i].sh_addr;
nrela = shdr[i].sh_size / sizeof(Elf_Rela);
for (j = 0; j < nrela; j++) {
COPYOUT(abase + j, &a, sizeof(a));
error = __elfN(reloc)(ef, __elfN(obj_symaddr),
&a, ELF_RELOC_RELA, base, off, val, len);
if (error != 0)
return (error);
}
break;
case SHT_REL:
rbase = (Elf_Rel *)(intptr_t)shdr[i].sh_addr;
nrel = shdr[i].sh_size / sizeof(Elf_Rel);
for (j = 0; j < nrel; j++) {
COPYOUT(rbase + j, &r, sizeof(r));
error = __elfN(reloc)(ef, __elfN(obj_symaddr),
&r, ELF_RELOC_REL, base, off, val, len);
if (error != 0)
return (error);
}
break;
}
}
return (0);
}
/* Look up the address of a specified symbol. */
static Elf_Addr
Make our ELF64 type definitions match standards. In particular this means: o Remove Elf64_Quarter, o Redefine Elf64_Half to be 16-bit, o Redefine Elf64_Word to be 32-bit, o Add Elf64_Xword and Elf64_Sxword for 64-bit entities, o Use Elf_Size in MI code to abstract the difference between Elf32_Word and Elf64_Word. o Add Elf_Ssize as the signed counterpart of Elf_Size. MFC after: 2 weeks
2005-12-18 04:52:37 +00:00
__elfN(obj_symaddr)(struct elf_file *ef, Elf_Size symidx)
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
{
Elf_Sym sym;
Elf_Addr base;
Enable warnings in EFI boot code Set WARNS if not set for EFI boot code and fix the issues highlighted by setting it. Most components are set to WARNS level 6 with few being left at lower levels due to the amount of changes needed to fix at higher levels. Error types fixed: * Missing / invalid casts * Missing inner structs * Unused vars * Missing static for internal only funcs * Missing prototypes * Alignment changes * Use of uninitialised vars * Unknown pragma (intrinsic) * Missing types etc due to missing includes * printf formatting types Reviewed by: emaste (in part) MFC after: 2 weeks X-MFC-With: r293268 Sponsored by: Multiplay Differential Revision: https://reviews.freebsd.org/D4839
2016-01-12 02:17:39 +00:00
if (symidx >= ef->e_shdr[ef->symtabindex].sh_size / sizeof(Elf_Sym))
Add the loader side of support for preloading ELF relocatable object format modules, which are currently only used on the amd64 platform. This initial implementation just parses enough of the module to allow it to extract dependencies and load all the bits into the right place in memory, so the kernel must still do the full relocation and linking. The details of the loaded sections are passed to the kernel by supplying a copy of the ELF section header table as module metadata with the MODINFOMD_SHDR tag.
2004-08-29 00:48:42 +00:00
return (0);
COPYOUT(ef->e_shdr[ef->symtabindex].sh_addr + symidx * sizeof(Elf_Sym),
&sym, sizeof(sym));
if (sym.st_shndx == SHN_UNDEF || sym.st_shndx >= ef->hdr.e_shnum)
return (0);
base = ef->e_shdr[sym.st_shndx].sh_addr;
if (base == 0)
return (0);
return (base + sym.st_value);
}
Reference in a new issue Copy permalink