Cache-as-Ram (no fill mode) Executable Code

Cache-as-Ram (no fill mode) Executable Code

By : user2951314
Date : November 18 2020, 01:01 AM
To fix this issue Coreboot originally used CAR to save C stack in L1 data cache: http://rere.qmqm.pl/~mirq/cache_as_ram_lb_09142006.pdf http://www.coreboot.org/images/6/6c/LBCar.pdf
To execute code, we should switch unified L2 to CAR mode, then L1i (you should know that most modern desktop/application CPUs has separated L1: one for data - L1d - with read/write and other - read-only L1i for code) will be able to read code from CAR L2. Such mode was implemented in "UBRX - Universal BIOS Recovery console for x86 PCs" (akeo): http://pete.akeo.ie/2011/08/ubrx-l2-cache-as-instruction-ram.html
code :

Share : facebook icon twitter icon
Does this code fill the CPU cache?

Does this code fill the CPU cache?

By : Jan Olsson
Date : March 29 2020, 07:55 AM
like below fixes the issue It depends on the actual code size - 10 lines of code can be little or much - and of course on the actual machine.
However, Method 2 violently violates this decades rule of thumb: instructions are cheap, memory access is not.
How to turn on emacs auto-fill-mode only for code comments?

How to turn on emacs auto-fill-mode only for code comments?

By : Netquart Dev
Date : March 29 2020, 07:55 AM
will help you If you want Emacs to auto-fill comments you must not make comment-auto-fill-only-comments a local variable:
code :
(setq comment-auto-fill-only-comments t)
(add-hook 'ruby-mode-hook 
          (lambda () ((set (make-local-variable 'comment-auto-fill-only-comments) t)))
(add-hook 'text-mode-hook 
          (lambda () (auto-fill-mode -1)))
In emacs Python mode, how do I set a different auto-fill width for docstrings and code?

In emacs Python mode, how do I set a different auto-fill width for docstrings and code?

By : James Roberts
Date : March 29 2020, 07:55 AM
I wish did fix the issue. I would like to have auto-fill set to 79 columns for code sections and 72 for docstrings to get automatic PEP8 compliance. There seems to be an option to do this for Lisp mode (emacs-lisp-docstring-fill-column) but not for Python. , Current python-mode.el provides
code :
(defcustom py-docstring-fill-column 72 [...]

(defcustom py-comment-fill-column 79 [...]
LD errors while linking 16-bit real mode code into a Multiboot compliant ELF executable

LD errors while linking 16-bit real mode code into a Multiboot compliant ELF executable

By : Swetanka Jaiswal
Date : March 29 2020, 07:55 AM
Any of those help I'm writing a Multiboot compliant ELF executable containing my 32-bit kernel. My primary problem is that I'm receiving a series of linker errors while producing my executable: , Reason for the Linker Error
This error you received:
code :
0000004a <realmode1>:

[bits 16]
mov ax,[vid_mode] ; Linker error
  63:   a1 0a 00                mov    ax,ds:0xa
                        64: R_386_16    .text
mov cx,[vid_mode] ; Linker error
  9a:   8b 0e 0a 00             mov    cx,WORD PTR ds:0xa
                        9c: R_386_16    .text
mov bx,[vid_mode] ; ; Linker error
  b2:   8b 1e 0a 00             mov    bx,WORD PTR ds:0xa
                        b4: R_386_16    .text
jmp 0x8:pm1 ; Linker error
  c5:   ea ca 00 08 00          jmp    0x8:0xca
                        c6: R_386_16    .text
. = 100000;
.text : { *(.text) }
.data : { *(.data) }
.bss  : { *(.bss)  }
mov ax,[vid_mode]
mov cx,[vid_mode]
mov bx,[vid_mode]
jmp 0x8:pm1
mov ax,[dword vid_mode]
mov cx,[dword vid_mode]
mov bx,[dword vid_mode]
jmp dword 0x8:pm1
mov ax,[dword vid_mode] ; Linker error
  63:   67 a1 0a 00 00 00       addr32 mov ax,ds:0xa
                        65: R_386_32    .text
mov cx,[dword vid_mode] ; Linker error
  9d:   67 8b 0d 0a 00 00 00    addr32 mov cx,WORD PTR ds:0xa
                        a0: R_386_32    .text
mov bx,[dword vid_mode] ; ; Linker error
  b8:   67 8b 1d 0a 00 00 00    addr32 mov bx,WORD PTR ds:0xa
                        bb: R_386_32    .text
jmp dword 0x8:pm1 ; Linker error
  ce:   66 ea d6 00 00 00 08    jmp    0x8:0xd6
  d5:   00
                        d0: R_386_32    .text

/* Multiboot spec uses 0x00100000 as a base */
PHYS_BASE = 0x00100000;
REAL_BASE = 0x00001000;

    . = PHYS_BASE;

    /* Place the multiboot record first */
    .multiboot : {

    /* This is the tricky part. The LMA (load memory address) is the
     * memory location the code/data is read into memory by the
     * multiboot loader. The LMA is after the colon. We want to tell
     * the linker that the code/data in this section was loaded into
     * RAM in the memory area above 0x100000. On the other hand the
     * VMA (virtual memory address) specified before the colon acts
     * like an ORG directive. The VMA tells the linker to resolve all
     * subsequent code starting relative to the specified VMA. The
     * VMA in this case is REAL_BASE which we defined as 0x1000.
     * 0x1000 is 4KB page aligned (useful if you ever use paging) and
     * resides above the end of the interrupt table and the
     * BIOS Data Area (BDA)

    __physreal_diff = . - REAL_BASE;
    .realmode REAL_BASE : AT(ADDR(.realmode) + __physreal_diff) {
        /* The __realmode* values can be used by code to copy
         * the code/data from where it was placed in RAM by the
         * multiboot loader into lower memory at REAL_BASE
         * . (period) is the current VMA */
        __realmode_vma_start = .;

        /* LOADADDR is the LMA of the specified section */
        __realmode_lma_start = LOADADDR(.realmode);
    . = ALIGN(4);
    __realmode_vma_end = .;
    __realmode_secsize   = ((__realmode_vma_end)-(__realmode_vma_start));
    __realmode_secsize_l = __realmode_secsize>>2;
    __realmode_lma_end   = __realmode_vma_start + __physreal_diff + __realmode_secsize;

    /* . (period) is the current VMA. We set it to the value that would
     * have been generated had we not changed the VMA in the previous
     * section. The .text section also specified the LMA = VMA with
     * AT(ADDR(.text))
    . += __physreal_diff;
    .text ALIGN(4K): AT(ADDR(.text)) {

    /* From this point the linker script is typical */
    .data ALIGN(4K) : {

    .data ALIGN(4K) : {

    /* We want to avoid this section being placed in low memory */
    .eh_frame : {

    .bss ALIGN(4K): {

    /* The .note.gnu.build-id section will usually be placed at the beginning
     * of the ELF object. We discard it (if it is present) so that the
     * multiboot header is placed as early as possible in the file. The
     * multiboot header must appear in the first 8K and be on a 4 byte
     * aligned offset per the multiboot spec.
    /DISCARD/ : {
CODE32SEL equ 0x08
DATA32SEL equ 0x10
CODE16SEL equ 0x18
DATA16SEL equ 0x20
; Video driver code - switches the CPU back into real mode
; Then executes an int 0x10 instruction

%include "gdt.inc"

global do_vbe

bits 16
section .data.realmode
save_idt: dw 0
          dd 0
save_esp: dd 0
vid_mode: dw 0
real_ivt: dw (256 * 4) - 1      ; Realmode IVT has 256 CS:IP pairs
          dd 0                  ; Realmode IVT physical address at address 0x00000

align 4
mode_info:TIMES 129 dw 0        ; Buffer to store mode info from Int 10h/ax=4f01h
                                ; Plus additional bytes for the return status byte
                                ; at beginning of buffer

bits 32
section .text
    mov ax, [esp+4]             ; Retrieve videomode passed on stack
    pushad                      ; Save all the registers
    pushfd                      ; Save the flags (including Interrupt flag)
    mov word [vid_mode],ax
    mov [save_esp],esp
    sidt [save_idt]
    lidt [real_ivt]             ; We use a real ivt that points to the
                                ; physical address 0x00000 at the bottom of
                                ; memory. The IVT in real mode is 256*4 bytes
                                ; and runs from physical address 0x00000 to
                                ; 0x00400
    jmp CODE16SEL:pmode16

bits 16
section .text.realmode
    mov ax,DATA16SEL
    mov ds,ax
    mov es,ax
    mov fs,ax
    mov gs,ax
    mov ss,ax
    mov eax,cr0
    dec eax
    mov cr0,eax
    jmp 0:realmode1

    ; Sets real mode stack to grow down from 0x1000:0xFFFF
    mov ax,0x1000
    mov ss,ax
    xor sp,sp

    xor ax,ax
    mov ds,ax
    mov es,ax
    mov fs,ax
    mov gs,ax

    ; first zero out the 258 byte memory for the return function from getmodeinfo
    mov cx,(258/2)              ; 128 words + 1 word for the status return byte
    mov di,mode_info
    rep stosw

    mov ax,[vid_mode]
    xor ax,0x13
    jnz svga_mode

    ; Just a regular mode13
    mov ax,0x13
    int 0x10

    ; Fake a video mode structure with the stuff the kernel actually uses
    mov di, mode_info
    mov word [di+0x01],0xDD     ; mode attribs
    mov word [di+0x13],320      ; width
    mov word [di+0x15],200      ; height
    mov byte [di+0x1a],8        ; bpp
    mov byte [di+0x1c],1        ; memory model type = CGA
    mov dword [di+0x29],0xa0000 ; screen memory
    jmp done

    mov ax,0x4f01               ; Get mode info function
    mov cx,[vid_mode]
    or cx,0x4000                ; always try to use linear buffer
    mov di,mode_info+0x01
    int 0x10
    mov [mode_info],ah
    or ah,ah
    jnz done

    mov ax,0x4f02               ; Now actually set the mode
    mov bx,[vid_mode]
    or bx,0x4000
    int 0x10

    mov eax,cr0
    inc eax
    mov cr0,eax
    jmp dword CODE32SEL:pm1     ; To FAR JMP to address > 0xFFFF we need
                                ; to specify DWORD to allow a 32-bit address
                                ; in the offset portion. When this JMP is
                                ; complete CS will be CODE32SEL and processor
                                ; will be in 32-bit protected mode

bits 32
section .text
    mov eax,DATA32SEL
    mov ds,ax
    mov es,ax
    mov fs,ax
    mov gs,ax
    mov ss,ax
    mov dword esp,[save_esp]
    lidt [save_idt]
    popfd                       ; Restore flags (including Interrupt flag)
    popad                       ; Restore registers

    mov eax, mode_info          ; Return pointer to mode_info structure
#ifndef VESADRV_H
#define VESADRV_H

#include <stdint.h>

extern struct mode_info_t * do_vbe (const uint8_t video_mode);

struct mode_info_t {
    uint8_t status; /* Return value from Int 10/ax=4f01 */

    /* Rest of structure from OSDev Wiki
    uint16_t attributes;
    uint8_t winA,winB;
    uint16_t granularity;
    uint16_t winsize;
    uint16_t segmentA, segmentB;

    /* Real mode FAR Pointer.  Physical address
     * computed as (segment<<4)+offset
    uint16_t realFctPtr_offset; /* FAR Pointer offset */
    uint16_t realFctPtr_segment;/* FAR Pointer segment */

    uint16_t pitch; /* bytes per scanline */

    uint16_t Xres, Yres;
    uint8_t Wchar, Ychar, planes, bpp, banks;
    uint8_t memory_model, bank_size, image_pages;
    uint8_t reserved0;

    uint8_t red_mask, red_position;
    uint8_t green_mask, green_position;
    uint8_t blue_mask, blue_position;
    uint8_t rsv_mask, rsv_position;
    uint8_t directcolor_attributes;

    volatile void * physbase;  /* LFB (Linear Framebuffer) address */
    uint32_t reserved1;
    uint16_t reserved2;
} __attribute__((packed));

#ifndef GDT_H
#define GDT_H

#include <stdint.h>
#include <stdbool.h>

typedef struct
        unsigned short limit_low;
        unsigned short base_low;
        unsigned char base_middle;
        unsigned char access;
        unsigned char flags;
        unsigned char base_high;
} __attribute__((packed)) gdt_desc_t;

typedef struct {
    uint16_t limit;
    gdt_desc_t *gdt;
} __attribute__((packed)) gdtr_t;

extern void gdt_set_gate(gdt_desc_t gdt[], const int num, const uint32_t base,
                         const uint32_t limit, const uint8_t access,
                         const uint8_t flags);

static inline void gdt_load(gdtr_t * const gdtr, const uint16_t codesel,
                            const uint16_t datasel, const bool flush)
    /* Dummy variable used as a fake dependency to avoid optimization
     * reordering of the inline assembly. The flush (if requested) must
     * always come after we set the GDT, not before */
    int dummy;

    /* load the GDT register */
    __asm__ __volatile__ ("lgdt %[gdtr]"
                          : "=X"(dummy)
                          : [gdtr]"m"(*gdtr),
                          /* Dummy constraint to ensure what gdtr->gdt points at is fully
                           * realized into memory before we issue LGDT instruction */
                            "m"(*(const gdt_desc_t (*)[]) gdtr->gdt));

    /* This flushes the selector registers to ensure the new
     * descriptors are used. */
    if (flush) {
        /* The indirect absolute jump is because we can't
         * assume that codesel is an immediate value
         * as it may be passed in a register. We build a
         * far pointer in memory and indirectly jump through
         * that pointer. This explicitly sets CS selector */
        __asm__  __volatile__ (
                 "pushl %[codesel]\n\t"
                 "pushl $1f\n\t"
                 "ljmpl *(%%esp)\n"
                 "add $8, %%esp\n\t"
                 "mov %[datasel], %%ds\n\t"
                 "mov %[datasel], %%es\n\t"
                 "mov %[datasel], %%ss\n\t"
                 "mov %[datasel], %%fs\n\t"
                 "mov %[datasel], %%gs"
                 : /* No outputs */
                 : "X"(dummy),
#include "gdt.h"

/* Setup a descriptor in the Global Descriptor Table */
void gdt_set_gate(gdt_desc_t gdt[], const int num, const uint32_t base,
                  const uint32_t limit, const uint8_t access,
                  const uint8_t flags)
        /* Setup the descriptor base access */
        gdt[num].base_low = (base & 0xFFFF);
        gdt[num].base_middle = (base >> 16) & 0xFF;
        gdt[num].base_high = (base >> 24) & 0xFF;

        /* Setup the descriptor limits */
        gdt[num].limit_low = (limit & 0xFFFF);
        gdt[num].flags = ((limit >> 16) & 0x0F);

        /* Finally, set up the flags and access byte */
        gdt[num].flags |= (flags << 4);
        gdt[num].access = access;
%include "gdt.inc"

STACKSIZE equ 0x4000

bits 32
global mbentry

extern kmain

; Multiboot Header
section .multiboot
MBALIGN     equ 1<<0
MEMINFO     equ 1<<1
VIDINFO     equ 0<<2
MAGIC       equ 0x1BADB002

    dd MAGIC
    dd FLAGS

section .text
    mov esp, stack_top

    ; EAX = magic number. Should be 0x2badb002
    ; EBX = pointer to multiboot_info
    ; Pass as parameters right to left
    push eax
    push ebx
    call kmain

    ; Infinite loop to end program
    jmp endloop

section .bss
align 32
    resb STACKSIZE
#include <stdint.h>
#include <stdbool.h>
#include "vesadrv.h"
#include "gdt.h"

#define CODE32SEL 0x08
#define DATA32SEL 0x10
#define CODE16SEL 0x18
#define DATA16SEL 0x20

/* You can get this structure from GRUB's multiboot.h if needed
 * https://www.gnu.org/software/grub/manual/multiboot/html_node/multiboot_002eh.html
struct multiboot_info;

/* Values made available by the linker script */
extern void *__realmode_lma_start;
extern void *__realmode_lma_end;
extern void *__realmode_vma_start;

/* Pointer to graphics memory.Mark as volatile since
 * video memory is memory mapped IO. Certain optimization
 * should not be performed. */
volatile uint32_t * video_gfx_ptr;

/* GDT descriptor table */
gdt_desc_t gdt[NUM_GDT_ENTRIES];

/* Copy the code and data in the realmode section down into the lower
 * 64kb of memory @ 0x00001000. */
static void realmode_setup (void)
    /* Each of these __realmode* values is generated by the linker script */
    uint32_t *src_addr = (uint32_t *)&__realmode_lma_start;
    uint32_t *dst_addr = (uint32_t *)&__realmode_vma_start;
    uint32_t *src_end  = (uint32_t *)&__realmode_lma_end;

    /* Copy a DWORD at a time from source to destination */
    while (src_addr < src_end)
        *dst_addr++ = *src_addr++;

void gdt_setup (gdt_desc_t gdt[], const int numdesc)
    gdtr_t gdtr = { sizeof(gdt_desc_t)*numdesc-1, gdt };

    /* Null descriptor */
    gdt_set_gate(gdt, 0, 0x00000000, 0x00000000, 0x00, 0x0);
    /* 32-bit Code descriptor, flat 4gb */
    gdt_set_gate(gdt, 1, 0x00000000, 0xffffffff, 0x9A, 0xC);
    /* 32-bit Data descriptor, flat 4gb */
    gdt_set_gate(gdt, 2, 0x00000000, 0xffffffff, 0x92, 0xC);
    /* 16-bit Code descriptor, limit 0xffff bytes */
    gdt_set_gate(gdt, 3, 0x00000000, 0x0000ffff, 0x9A, 0x0);
    /* 16-bit Data descriptor, limit 0xffffffff bytes */
    gdt_set_gate(gdt, 4, 0x00000000, 0xffffffff, 0x92, 0x8);

    /* Load global decriptor table, and flush the selectors */
    gdt_load(&gdtr, CODE32SEL, DATA32SEL, true);

int kmain(struct multiboot_info *mb_info, const uint32_t magicnum)
    struct mode_info_t *pMI;
    uint32_t pixel_colors = 0;

    /* Quiet compiler about unused variables */
    (void) mb_info;
    (void) magicnum;

    /* Setup the GDT */
    gdt_setup(gdt, NUM_GDT_ENTRIES);

    /* Setup real mode code and data */

    /* Switch to video mode 0x13 (320x200x256)
     * The physical address of the mode 13 video memory is
     * 0xa0000 */
    pMI = do_vbe(0x13);
    video_gfx_ptr = pMI->physbase;

    /* Display part of the VGA palette as a test pattern */
    for (int pixelpos = 0; pixelpos < (320*200); pixelpos++) {
        if ((pixel_colors & 0xff) == (320/4))
            pixel_colors = 0;
        pixel_colors += 0x01010101;
        video_gfx_ptr[pixelpos] = pixel_colors;
    return 0;
nasm -f elf32 -g -F dwarf -o multiboot.o multiboot.asm
nasm -f elf32 -g -F dwarf -o vesadrv.o vesadrv.asm
i686-elf-gcc -std=c99 -g -m32 -O3 -c -fno-exceptions -nostdlib -ffreestanding -Wall -Wextra -o kernel.o kernel.c
i686-elf-gcc -std=c99 -g -m32 -O3 -c -fno-exceptions -nostdlib -ffreestanding -Wall -Wextra -pedantic -o gdt.o gdt.c -lgcc
i686-elf-gcc -m32 -Tlink.ld -ffreestanding -nostdlib -o multiboot.elf multiboot.o kernel.o gdt.o vesadrv.o -lgcc
WinError 6 - The handle is invalid with executable but code works on debug mode

WinError 6 - The handle is invalid with executable but code works on debug mode

By : user3514264
Date : March 29 2020, 07:55 AM
will help you I'm using a software that handle video processing in a powershell subprocess. , The problem is with psa1_path variable
code :
C:/Users/${USERNAME}\projects\demo\cmd\powershell.ps1 -m -v 'CC 2018' -wait windowstyle hidden
p = subprocess.Popen(['C:\\WINDOWS\\system32\\WindowsPowerShell\\v1.0\\powershell.exe',
                             stdout=f, shell=True, stdin=subprocess.DEVNULL)
Related Posts Related Posts :
  • Property 'throw' does not exist on type 'typeof Observable'
  • Unable to deploy SAPUI5 application with 404 error
  • Xcode 10 build fail rm permission denied in /bin/sh/
  • Handling quorum writies fail on Cassandra
  • How to finalize log file just after time is over when using logback SizeAndTimeBasedFNATP?
  • Write custom widget with GTK3
  • How to remove UINavigationBar inner shadow in iOS 7?
  • Wordpress dynamic widget by location?
  • XDocument replace all node value with lower case
  • Invalid tagdir attribute while web-fragment is used
  • Set default font/text size in RTF Control
  • what video formats supported by chromecast device (mp4,flv,m3u8,flv,avi)?
  • sbcl - how to muffle "undefined variable" warning?
  • Get an eventbrite event ID
  • Obtain date without timestamp in DB2
  • Cron job to SFTP files in a directory
  • Draw two head arrows in fabric.js
  • YOLO darknet vs darkflow
  • Parse custom rss tags using Rome API
  • Creating text editor like EDIT on Command Prompt using FreePascal
  • Subplot hides xlabel
  • NServiceBus long running process that timesout
  • Qt4 - How to add a row in a QTableWidget when sorting is enable?
  • Cancel command in Grunt that hasn't been launched
  • Canvas globalCompositeOperation is not working correctly
  • HTML.Kendo().Dropdownlist set default item
  • vagrant up failed, /dev/vboxnetctl: no such file or directory
  • How to use jssor carousel to build client carousal
  • Theory what is meant by validating security concerns for these user groups
  • How to build correct SPARQL Query
  • Connecting to MySQL (on Google Cloud SQL) via JDBC and IPv6?
  • Fetching transaction in several iterations
  • is it possible to use dropbox datastore api as shared datastore?
  • phpExcel 1.8 Named Ranges in Formula
  • Cisco VPN IP address
  • No address associated with hostname using vagrant
  • 2 way data-binding between a polymer component and a model?
  • How can I rename an already published app in the Windows Dev Center?
  • Using Chef 12 , Chef Client unable to connect to chef Server
  • GCM notification not working (GoogleCloudMessaging)
  • Measured GFLOPS is greater then theoretical GFLOPS
  • Progress ABL - strip and add to temp table
  • Errors occuring when make-ing Aircrack-ng on raspberry pi
  • Detecting Handedness from Device Use
  • Query a manual list of data items
  • Similar to pivot - Windows 8.1 Desktop app
  • CKEDITOR And Styles configuration with only toolbar
  • Segmentation fault due to data alignment issue on MIC
  • how to connect a Java app ( java code ) to cloudant?
  • How to download a file/folder from remote (openshift) to local system
  • Where can i find customeditors in JBPM 6
  • Pig get distinct rows with counts
  • In Flex when is yylineno updated?
  • How does a semantic reasoner for protegé is made?
  • how to stop another rows to be select after select a row in uitableview
  • Setup cmd.exe parameter when starting ConEmu
  • WPF how to get the same visual size in different devices
  • How to convert tiff to searchable pdf using alfresco and tesseact?
  • Intel Media SDK: Regarding mfxIMPL Usage
  • How to pass values to SOAP service using web service consumer?
  • shadow
    Privacy Policy - Terms - Contact Us © ourworld-yourmove.org