Its only purpose left was to flush buffers, which isn't necessary after
just open/read, so we can move that to dir_write().
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@1410 6ac86273-5f31-0410-b378-82cca8765d1b
dos_setfsize still needed to merge changes for SHARE so the call was replaced
by a new dos_merge_file_changes() call.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@1405 6ac86273-5f31-0410-b378-82cca8765d1b
you reboot while files are open... Reported by Hardi Stengelin :-).
Always test *link_fat* result. New function is_free_cluster. Extra
checks in *link_fat* (bad r/w offset, i/o, write) and *next_cluster*
(dangling chain: bad value in chain) with short messages (no msg for
bad chain start). Shorter "Bad DPB" (FAT size) msg. More comments!
Update FS INFO on disk only when needed. Check if chain EOF was at
expected place when extending. Call *link_fat* before! *setdstart*
when a FAT chain grows. [Q: Does 0 byte write always trunc? Why?]
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@1358 6ac86273-5f31-0410-b378-82cca8765d1b
"HLT on idle" (dosidle.asm, int2f.asm) and KEYBUF=n[,m] to relocate
the keyboard buffer to another area in the 40:ac..40:1ff range. Global
functions and variables: proto.h DosIdle_hlt, init-mod.h and globals.h
DosIdle_hlt() and HaltCpuWhileIdle.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@1339 6ac86273-5f31-0410-b378-82cca8765d1b
Change prototype of DosGetFree so it returns AX (Arkady)
Checking the return value for FFFF is enough to check for failure.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@1008 6ac86273-5f31-0410-b378-82cca8765d1b
and let it return a long instead of using a *err pointer.
Call the resulting function NETWORK_REDIRECTOR_MX (multiplex)
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@893 6ac86273-5f31-0410-b378-82cca8765d1b
Seems to be more consistent now. read_char_sft_dev handles ^C at a fairly
high level. So raw_get_char doesn't need to worry about sft's anymore.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@885 6ac86273-5f31-0410-b378-82cca8765d1b
- ^C didn't work for int21/ah=3f. Corrected -- Do_DosIdle_loop still needs
to wait even if it doesn't call the idle int
- ^C needs to be echoed to STDOUT for low character functions (ah<0xd) but
to the device that is read from for handle functions. So we need to pass
either the STDOUT sft or that device on to the ^C echoer.
- The ^C echo really should print ^C\r\n.
- misc cleanups in break.c.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@883 6ac86273-5f31-0410-b378-82cca8765d1b
them pop the stack (smaller code than using bx).
Tell Watcom which registers are clobbered; save some more registers
for intr() -- especially for intr() this helps.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@849 6ac86273-5f31-0410-b378-82cca8765d1b
all other compilers can use it with pascal or "register" calling conventions.
Saves ~50 bytes for the init code.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@845 6ac86273-5f31-0410-b378-82cca8765d1b
Internally the kernel uses two near fnodes though, to save on codesize
and fmemcpy's if necessary. Having memory management on two fnodes is
a little silly but I just want to make sure with the panic message
that we never accidentally try to use three near fnodes at the same time.
(two are used at the same time by rename, commit, and merge_file_changes).
This can be cleaned up later.
Anyway. 644736 bytes free conv memory isn't bad.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@821 6ac86273-5f31-0410-b378-82cca8765d1b
FAT16 partitions (just like FAT16 kernels do)
* added extra checks to make sure that invalid FAT entries are never
followed
* made put_console() public to be able to print a message in case of
FAT corruption
* some small optimizations and header cleanups (some suggested by Arkady)
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@753 6ac86273-5f31-0410-b378-82cca8765d1b
Fix int2f/ax=1216 (needs sft handles, not psp handles, and to return
the relative handle in BX, and to deal with carry)
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@712 6ac86273-5f31-0410-b378-82cca8765d1b
This solves problems with int24 handlers that call int21/ah=59.
Let truename and blockrw return with an error if there is a critical error.
Don't overwrite the critical error # with the DOS error #.
git-svn-id: https://svn.code.sf.net/p/freedos/svn/kernel/trunk@709 6ac86273-5f31-0410-b378-82cca8765d1b