[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Inquiring cttys
Julian Reschke writes:
> A while back, someone (Juergen?) posted a new Fcntl for u:\proc to inquire
> the controlling tty of a process. Please repost that one, because it would
> be very useful for my ps.ttp...
ok you want a patch... :)
Index: file.h
@@ -391,6 +391,7 @@
#define PTRACESTEP (('P'<< 8) | 10)
#define PTRACE11 (('P'<< 8) | 11)
#define PLOADINFO (('P'<< 8) | 12)
+#define PFSTAT (('P'<< 8) | 13)
struct ploadinfo {
/* passed */
Index: procfs.c
@@ -561,11 +561,23 @@
*((long *)buf) = (long)&p->magic;
return 0;
case PBASEADDR:
- *((long *)buf) = (long)p->base;
+ if (p == rootproc)
+ *((long *)buf) = (long)_base;
+ else
+ *((long *)buf) = (long)p->base;
return 0;
case PCTXTSIZE:
*((long *)buf) = sizeof(CONTEXT);
return 0;
+ case PFSTAT:
+ {
+ FILEPTR *pf;
+ int pfd = (*(ushort *)buf);
+ if (pfd < MIN_HANDLE || pfd >= MAX_OPEN ||
+ !(pf = p->handle[pfd]))
+ return EIHNDL;
+ return (*pf->fc.fs->getxattr)(&pf->fc, (XATTR *)buf);
+ }
case PSETFLAGS:
{
int newflags = (ushort)(*(long *)buf);
PFSTAT does FSTAT on processes fd passed in first word of stat buffer,
to get the control tty use -1 and then search the `inode' in /dev or
/pipe like in ttyname(),
volatile struct stat sb;
*(short *)&sb = fd; /* -1 for /dev/tty */
if (r = Fcntl(procf, &sb, PFSTAT))
error ...
find device...
(volatile seems necessary because of the gcc Fcntl binding(?), can this be
fixed?)
i also changed PBASEADDR to return the kernels real basepage even
when it unlinked itself out of the chain (see main.c), that should help
stephen read kernel memory easier etc. (btw your ps doesn't use it now
(ADDR), right? :)
later...
Juergen
--
J"urgen Lock / nox@jelal.north.de / UUCP: ..!uunet!unido!uniol!jelal!nox
...ohne Gewehr
PGP public key fingerprint = 8A 18 58 54 03 7B FC 12 1F 8B 63 C7 19 27 CF DA