OSSP CVS Repository
OSSP CVS Repository |
|
|
| Date | Version | Description |
|---|---|---|
| 1.41 | Check-in
[2970]:
moved into its own al project
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.40 | Check-in
[2933]:
get rid of superflous unsigned<0 tests
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.39 | Check-in
[2922]:
make al_setlabel conditional (use oldlabel == NULL for unconditional
behavíour)
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.38 | Check-in
[2918]:
new al_spanlabel method
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.37 | Check-in
[2916]:
new al_setlabel method to modify the data label within an
assembly line.
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.36 | Check-in
[2851]:
buffers can move to different assembly lines so that the freemem
function can reference a dangling al pointer to reach bfree().
We now reference the bfree() function directly as this is supposed
to remain valid even after an assembly line has been freed.
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.35 | Check-in [2821]: remove remaining GCC warnings By rse. (diff) | |
| 1.34 | Check-in
[2730]:
size_t is unsigned -> get rid of possible negative offsets that weren't used
anywhere. remove sanity checks for 'unsigned < 0'.
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.33 | Check-in
[2727]:
prepending to a zero length chunk is now possible, the chunk
is aligned to the end of the buffer
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.32 | Check-in
[2667]:
optimize al_splice to unlink/relink several chunks in a single step
added necessary list macros to list.h
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.31 | Check-in
[2653]:
new span traversal mode
add direction to al_flatten to support span traversal
utility method al_firstlabel
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.30 | Check-in
[2651]:
seeking backwards to EOF position is obviously ok as well
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.29 | Check-in
[2649]:
significant API change.
now support data labels.
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.28 | Check-in
[2648]:
prepend used wrong macros (cut&paste error)
traverse_next didn't compute step correctly and didn't clear skip
from initial seek when moving to next/previous chunk
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.27 | Check-in
[2622]:
beautification
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.26 | Check-in
[2620]:
correctly initialize callback userdata
get rid of debug printf
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.25 | Check-in
[2619]:
after splitting initial chunk, the split offset ("skip") needs
to be reset to zero.
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.24 | Check-in
[2618]:
change in al_attach_buffer API
back out rse's change to al_traverse parameters. parameter order is: object ptr(if any), input parameters, output parameters PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.23 | Check-in [2616]: Undocumented internal code is usually always ok (at least as long as it is self-describing enough ;-), but undocumented internal data is most of the time nasty. It especially makes it hard to review anything in-depth because only the author has all of the gory details in mind all the time. Hence internal data should be documented at least a little bit to make clear for what each structure member is for. So, here it is. Feel free to fix my quick descriptions. By rse. (diff) | |
| 1.22 | Check-in
[2613]:
Be pedantic about consistent argument orders and move al_tx_t argument
to second position where it is for all other al_tx_t based functions:
-al_traverse(al_t *al, size_t off, size_t n, al_td_t dir, al_tx_t *tx) +al_traverse(al_t *al, al_tx_t *tx, size_t off, size_t n, al_td_t dir) By rse. (diff) | |
| 1.21 | Check-in
[2611]:
Before diving into the semantics, bring syntax into standard
OSSP shape by applying the following cosmetics:
- add newlines between "if" and "return" at some statements - make sure indentation is exactly 4 spaces (were 8 at dispose_chunk, etc) - consistently use no braces for "if" or "else" clauses if only single statement exists in body. - we do not indicate the not-used-return-code via (void) on statements, because both every compiler and human is smart enough to see the intention. THERE ARE NO FUNCTIONAL CHANGES IN THIS COMMIT AT ALL. By rse. (diff) | |
| 1.20 | Check-in
[2609]:
added al_traverse_end to API, to allow future cleanup/locking
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.19 | Check-in
[2608]:
moved generic list macros to separate header
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.18 | Check-in [2606]: Ok, as arranged with Michael: use 'assembly line' as the official long name By rse. (diff) | |
| 1.17 | Check-in [2603]: decouple pos/neg. offset handling from search direction decision By thl. (diff) | |
| 1.16 | Check-in [2600]: using REMTAIL ... closer approximity to the CPU By thl. (diff) | |
| 1.15 | Check-in
[2599]:
make comment clearer about not resizing chunks in a shared buffer
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.14 | Check-in [2596]: add the usual amount of license fun By rse. (diff) | |
| 1.13 | Check-in
[2594]:
use default signedness in API and internally
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.12 | Check-in
[2591]:
code cleanup
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.11 | Check-in
[2590]:
optimization for short sizes was buggy.
clarify comments
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.10 | Check-in
[2589]:
move private al_buffer_t declaration to .c file
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.9 | Check-in
[2580]:
EX support
more comments
insert failure path moved to top of al_splice
chunk copy failure path leaves data structures intact ?!
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.8 | Check-in
[2579]:
remove debugging assert()
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.7 | Check-in
[2578]:
al_copy function, like al_flatten but into a target assembly list
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.6 | Check-in
[2577]:
INSERT/REMOVE didn't maintain list header
test code
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.5 | Check-in
[2576]:
stop traversal at span boundary
code cleanup
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.4 | Check-in
[2575]:
code cleanup
get rid of AL_ALL_BYTES, not very useful
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.3 | Check-in
[2574]:
splice operation
traversal needs alloc/free function for context
lost of comments
PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.2 | Check-in [2573]: added memory allocation policy needs to pass through context to lower functions PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. (diff) | |
| 1.1 | Check-in [2571]: first iteration through buffer data type PR: Submitted by: Reviewed by: Approved by: Obtained from: By mlelstv. |