Page 1 of 2
apply a structure to memory data
Posted: Mon Jun 04, 2007 10:04 pm
by frederic
Hello, i don't know if this pb is linux specific but i would like to apply a structure to data memory.
The function uname_ return a pointer to some structured information and i want to work with this info easily with the structure "utsname" but how ?
For the moment, i use PeekS, but it stops at the end of first string (normal it's his job).
Code: Select all
Structure utsname
sysname.s
nodename.s
release.s
version.s
machine.s
EndStructure
Global *test=AllocateMemory(SizeOf(utsname))
If uname_(*test) = 0
Debug PeekS(*test)
Else
Debug "error"
EndIf
Posted: Mon Jun 04, 2007 10:08 pm
by Kaeru Gaman
use other components than .s
.s is a 4byte long pointer to a string, not the string itself.
Posted: Mon Jun 04, 2007 10:10 pm
by netmaestro
I'm not 100% sure what it is you're asking, but this is how you might set/retrieve the strings in the structure:
Code: Select all
Structure utsname
sysname.s
nodename.s
release.s
version.s
machine.s
EndStructure
Global *test.utsname = AllocateMemory(SizeOf(utsname))
With *test
\sysname = "MySystem"
\nodename = "MyNode"
\release = "1.0"
\version = "1"
\machine = "MyMachine"
EndWith
With *test
Debug \sysname
Debug \nodename
Debug \release
Debug \version
Debug \machine
EndWith
Posted: Mon Jun 04, 2007 10:10 pm
by srod
Code: Select all
Structure utsname
sysname.s
nodename.s
release.s
version.s
machine.s
EndStructure
Global *test.utsname=AllocateMemory(SizeOf(utsname))
*test\sysname = "Sys"
*test\nodename = "Node"
Debug *test\sysname
Beware though that when freeing the memory pointed to by *test, the memory allocated to the individual strings will not be freed.
There's a thread somewhere in these forums which discusses this and shows a workaround.
**EDIT: too slow by far!
Posted: Mon Jun 04, 2007 10:12 pm
by netmaestro
Ha! I did them all and still beatcha!
Posted: Mon Jun 04, 2007 10:14 pm
by srod
netmaestro wrote:Ha! I did them all and still beatcha!
It's this damn dialup internet connection I'm having to use at the moment!
That's my excuse anyhow!
Posted: Mon Jun 04, 2007 10:14 pm
by frederic
Your example doesn't works with the function uname_
I've an Invalid Memory Access on this line " Debug \sysname"
Code: Select all
Structure utsname
sysname.s
nodename.s
release.s
version.s
machine.s
EndStructure
Global *test.utsname=AllocateMemory(SizeOf(utsname))
uname_(*test)
With *test
Debug \sysname
Debug \nodename
Debug \release
Debug \version
Debug \machine
EndWith
Posted: Mon Jun 04, 2007 10:15 pm
by srod
What is uname_() ?
Posted: Mon Jun 04, 2007 10:17 pm
by netmaestro
The structure is probably wrong then, could they be fixedstrings or byte arrays? Something other than null-terminated strings, anyhow.
Posted: Mon Jun 04, 2007 10:17 pm
by frederic
uname_ is a linux function that retrieve some system information
Code: Select all
int uname(struct utsname *buf);
DESCRIPTION
uname returns system information in the structure pointed to by buf. The utsname struct is defined in <sys/utsname.h>:
struct utsname {
char sysname[];
char nodename[];
char release[];
char version[];
char machine[];
#ifdef _GNU_SOURCE
char domainname[];
#endif
};
The length of the arrays in a struct utsname is unspecified; the fields are NUL-terminated.
with this i've no error, but i don't know if it's correct
Code: Select all
Structure utsname
sysname.s{255}
nodename.s{255}
release.s{255}
version.s{255}
machine.s{255}
EndStructure
Posted: Mon Jun 04, 2007 10:18 pm
by srod
Indeed, we can now see that this is an entirely different kind of structure.
Thinking...
Posted: Mon Jun 04, 2007 10:19 pm
by Kale
another way:
Code: Select all
Structure utsname
sysname.s
nodename.s
release.s
version.s
machine.s
EndStructure
Procedure uname_()
Static MyStruct.utsname
MyStruct\sysname = "sysname"
MyStruct\nodename = "nodename"
MyStruct\release = "release"
MyStruct\version = "version"
MyStruct\machine = "machine"
ProcedureReturn @MyStruct
EndProcedure
Global *test.utsname = uname_()
Debug *test\sysname
Debug *test\nodename
Debug *test\release
Debug *test\version
Debug *test\machine
Posted: Mon Jun 04, 2007 10:25 pm
by frederic
ok with a scan of memory it appears that the data are (some datas
) :
Code: Select all
0807CA74 4C 69 6E 75 78 00 00 00 00 00 00 00 00 00 00 00 Linux...........
0807CA84 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CA94 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CAA4 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CAB4 00 50 75 72 65 66 72 69 65 6E 64 73 00 00 00 00 .Purefriends....
0807CAC4 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CAD4 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CAE4 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CAF4 00 00 32 2E 36 2E 32 30 2D 31 36 2D 67 65 6E 65 ..2.6.20-16-gene
0807CB04 72 69 63 00 00 00 00 00 00 00 00 00 00 00 00 00 ric.............
0807CB14 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
0807CB24 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................
so i must to count how many bytes there are for a string or "WE" know how many byte is a char xxxx[]; in linux language
Posted: Mon Jun 04, 2007 10:28 pm
by srod
Not sure you've followed the thread correctly kale. uname_() is an external function.
frederick, I'd be a little surprised if the fixed length strings work if the structure in question is supposed to contain variable length strings with a null terminator.
The thing is about PB structures, string fields are stored as pointers rather than an array of charcaters embedded within the structure. This explains the invalid memory access error etc.
Fixed length strings are moving in the right direction in that the characters are stored directly within the structure etc. but here we stumble because of the the fact that your structures needs variable length strings etc.
I reckon you'll just have to allocate a large buffer and write your own code to read and write the string fields etc. Should be easy enough.
Posted: Mon Jun 04, 2007 10:45 pm
by Fred
The first time ever i've seen a structure like that...