Merge chapter/process into 0096079

Author: SamyPesse

Chapter-8/README.md

@@ -1,4 +1,4 @@
-## Chapter 8: Memory management: physical and virtual
+## Chapter 8: Theory: physical and virtual memory
 
 In the chapter related to the GDT, we saw that using segmentation a physical memory address is calculated using a segment selector and an offset.
 
@@ -41,14 +41,14 @@ The two types of entries (table and directory) look like the same. Only the fiel
 * `A`: indicate if the page or table was accessed
 * `D`: (only for pages table) indicate if the page was written
 * `PS` (only for pages directory) indicate the size of pages:
-    * 0 = 4ko
-    * 1 = 4mo
+    * 0 = 4kb
+    * 1 = 4mb
 
-**Note:** Physical addresses in the pages diretcory or pages table are written using 20 bits because these addresses are aligned on 4ko, so the last 12bits should be equal to 0.
+**Note:** Physical addresses in the pages diretcory or pages table are written using 20 bits because these addresses are aligned on 4kb, so the last 12bits should be equal to 0.
 
 * A pages directory or pages table used 1024*4 = 4096 bytes = 4k
-* A pages table can address 1024 * 4k = 4 Mo
-* A pages directory can address 1024 * (1024 * 4k) = 4 Go
+* A pages table can address 1024 * 4k = 4 Mb
+* A pages directory can address 1024 * (1024 * 4k) = 4 Gb
 
 #### How to enable pagination?
 
@@ -63,5 +63,16 @@ asm("  mov %%cr0, %%eax; \
 
 But before, we need to initialize our pages directory with at least one pages table.
 
+#### Identity Mapping
+
+With the identity mapping model, the page will apply only to the kernel as the first 4 MB of virtual memory coincide with the first 4 MB of physical memory:
+
+![Identity Mapping](identitymapping.png)
+
+This model is simple: the first virtual memory page coincide to the first page in physical memory, the second page coincide to the second page on physical memory and so on ...
+
+
+
+
 
 

Chapter-8/identitymapping.png

@@ -1,13 +1,15 @@
-### Summary
+# Summary
 
+* [Introduction](README.md)
 * [Introduction about the x86 architecture and about our OS](Chapter-1/README.md)
 * [Setup the development environment](Chapter-2/README.md)
 * [First boot with GRUB](Chapter-3/README.md)
 * [Backbone of the OS and C++ runtime](Chapter-4/README.md)
 * [Base classes for managing x86 architecture](Chapter-5/README.md)
 * [GDT](Chapter-6/README.md)
 * [IDT and interrupts](Chapter-7/README.md)
-* [Memory management: physical and virtual](Chapter-8/README.md)
+* [Theory: physical and virtual memory](Chapter-8/README.md)
+* [Memory management: physical and virtual](chapter9/README.md)
 * Process management and multitasking
 * External program execution: ELF files
 * Userland and syscalls
@@ -18,4 +20,5 @@
 * EXT2 read-only filesystems
 * Standard C library (libC)
 * UNIX basic tools: sh, cat
-* Lua interpreter
+* Lua interpreter
+

SUMMARY.md

@@ -0,0 +1,34 @@
+# Memory management: physical and virtual
+
+The kernel knows the size of the physical memory available thanks to [GRUB](../Chapter-3/README.md).
+
+In our implementation, the first 8 megabytes of physical memory will be reserved for use by the kernel and will contain:
+
+- The kernel
+- GDT, IDT et TSS
+- Kernel Stack
+- Some space reserved to hardware (video memory, ...)
+- Page directory and pages table for the kernel
+
+The rest of the physical memory is freely available to the kernel and applications.
+
+![Physical Memory](physicalmemory.png)
+
+
+### Virtual Memory Mapping
+
+The address space between the beginning of memory and `0x40000000` address is the kernel space, while the space between the address `0x40000000` and the end of the memory corresponds to user space:
+
+![Virtual Memory](virtualmemory.png)
+
+The kernel space in virtual memory, which is using 1Gb of virtual memory, is common to all tasks (kernel and user).
+
+This is implemented by pointing the first 256 entries of the task page directory to the kernel page directory (In [vmm.cc](https://github.com/SamyPesse/How-to-Make-a-Computer-Operating-System/blob/master/src/kernel/arch/x86/vmm.cc#L204)):
+
+```cpp
+/* 
+ * Kernel Space. v_addr < USER_OFFSET are addressed by the kernel pages table
+ */
+for (i=0; i<256; i++) 
+    pdir[i] = pd0[i];
+```