Difference between revisions of "Talk:Kernel Syscalls"

From The iPhone Wiki
Jump to: navigation, search
 
(5 intermediate revisions by 4 users not shown)
Line 1: Line 1:
 
i dont know if its right, pls correct me if im wrong :) --[[User:Zmaster|Zmaster]] 13:58, 29 November 2011 (MST)
 
i dont know if its right, pls correct me if im wrong :) --[[User:Zmaster|Zmaster]] 13:58, 29 November 2011 (MST)
   
  +
<del>
 
HTTP/i0nic:
 
HTTP/i0nic:
 
The following is hardly "bogus".
 
The following is hardly "bogus".
Line 6: Line 7:
 
0x30d2ad58 <chown+4>: svc 0x00000080
 
0x30d2ad58 <chown+4>: svc 0x00000080
 
is a direct disassembly of libSystem. You can see that with gdb on a jb device.
 
is a direct disassembly of libSystem. You can see that with gdb on a jb device.
  +
</del>
   
SVC is an ARM instruction to invoke a "supervisor call". The 0x80 is the call #, because the chip allows an interrupt vector, much like Intel's INT instruction. Then, you place the syscall # (in the above example, chown) in r12. [[User:morpheus|morpheus]] ||3/1/2012, 20:01 EST.
+
<del>SVC is an ARM instruction to invoke a "supervisor call". The 0x80 is the call #, because the chip allows an interrupt vector, much like Intel's INT instruction. Then, you place the syscall # (in the above example, chown) in r12. [[User:morpheus|morpheus]] ||3/1/2012, 20:01 EST.
  +
</del>
  +
  +
-- Edit:
  +
  +
OOOOOH. Now I get it. He meant the CPU syscalls, not the kernel syscalls. This needs more research (who originally put that part?) {{unsigned|morpheus|2 Mar 2012 02:04 UTC}}
  +
:Look at page history. It seems like it came from [[User:ChronicDev|Chronic]]? Hardly imaginable he entered wrong infos. Maybe just very outdated? I have too few experience with ARM and kernel to contribute here. Feel free to remove wrong infos or just mention where it belongs to. --[[User:Http|http]] 01:14, 2 March 2012 (MST)
  +
  +
== OS thinko? ==
  +
  +
Morpheus: when you wrote "Most of these syscalls are the same as those of iOS" did you mean to write "Most of these syscalls are the same as those of OS X"? Weren't we already talking about iOS at the time? --[[User:Beej|beej]] 09:08, 20 March 2012 (MDT)
  +
  +
Beej: Yep. Serves me right for writing this when I'm so jetlagged. {{unsigned|Morpheus|17:29, March 20, 2012 (MST)}}
  +
  +
==Edit==
  +
Shouldn't we remove the sentance "A program to do so reliable on iOS has, in fact, been written, and produces the following output for iOS 6.0b1:" nearby the syscall list? It doesn't help, and goes against the fact that the list is for iOS 6 GM. --[[User:Awesomebing1|Awesomebing1]] ([[User talk:Awesomebing1|talk]]) 01:27, 2 October 2014 (UTC)

Latest revision as of 01:27, 2 October 2014

i dont know if its right, pls correct me if im wrong :) --Zmaster 13:58, 29 November 2011 (MST)

HTTP/i0nic: The following is hardly "bogus". 0x30d2ad54 <chown>: mov r12, #16  ; 0x10, being # of chown 0x30d2ad58 <chown+4>: svc 0x00000080 is a direct disassembly of libSystem. You can see that with gdb on a jb device.

SVC is an ARM instruction to invoke a "supervisor call". The 0x80 is the call #, because the chip allows an interrupt vector, much like Intel's INT instruction. Then, you place the syscall # (in the above example, chown) in r12. morpheus ||3/1/2012, 20:01 EST.

-- Edit:

OOOOOH. Now I get it. He meant the CPU syscalls, not the kernel syscalls. This needs more research (who originally put that part?) --The preceding unsigned comment was added by morpheus (talk) 2 Mar 2012 02:04 UTC. Please consult this page for more info on how to sign pages, and how to fix this.

Look at page history. It seems like it came from Chronic? Hardly imaginable he entered wrong infos. Maybe just very outdated? I have too few experience with ARM and kernel to contribute here. Feel free to remove wrong infos or just mention where it belongs to. --http 01:14, 2 March 2012 (MST)

OS thinko?

Morpheus: when you wrote "Most of these syscalls are the same as those of iOS" did you mean to write "Most of these syscalls are the same as those of OS X"? Weren't we already talking about iOS at the time? --beej 09:08, 20 March 2012 (MDT)

Beej: Yep. Serves me right for writing this when I'm so jetlagged. --The preceding unsigned comment was added by Morpheus (talk) 17:29, March 20, 2012 (MST). Please consult this page for more info on how to sign pages, and how to fix this.

Edit

Shouldn't we remove the sentance "A program to do so reliable on iOS has, in fact, been written, and produces the following output for iOS 6.0b1:" nearby the syscall list? It doesn't help, and goes against the fact that the list is for iOS 6 GM. --Awesomebing1 (talk) 01:27, 2 October 2014 (UTC)