Rev |
Age |
Author |
Path |
Log message |
Diff |
4539 |
5549 d 17 h |
jermar |
/trunk/uspace/srv/vfs/ |
Handle pending mounts using only one fibril. |
|
4538 |
5549 d 17 h |
svoboda |
/trunk/uspace/srv/console/ |
Revert clandestine removal of delayed cursor position update by c4456. |
|
4534 |
5551 d 17 h |
svoboda |
/trunk/uspace/srv/bd/file_bd/ |
Add forgotten files. |
|
4532 |
5551 d 23 h |
svoboda |
/trunk/uspace/srv/bd/ata_bd/ |
Write support in ata_bd. Not used, not tested. |
|
4531 |
5552 d 2 h |
svoboda |
/trunk/uspace/srv/bd/ata_bd/ |
Use named constants. |
|
4530 |
5552 d 3 h |
svoboda |
/trunk/uspace/srv/bd/ata_bd/ |
Move type declarations to a header file. |
|
4529 |
5552 d 14 h |
svoboda |
/trunk/uspace/srv/bd/ata_bd/ |
Properly support two disks per ATA controller. Fake (Qemu-only) check whether disk is present. Only create device nodes for detected disks. |
|
4528 |
5552 d 16 h |
svoboda |
/trunk/uspace/ |
Use 'fibril_local' modifier for fibril-local variables instead of '__thread'. Since it is decalred in fibril.h, it need not begin with '__'. |
|
4525 |
5554 d 14 h |
svoboda |
/trunk/uspace/srv/fb/ |
Fix coordinate issues in EGA fb. |
|
4524 |
5554 d 17 h |
decky |
/trunk/uspace/srv/devmap/ |
get back fine-grained locking (which was removed in commit 4458), but this time with proper fibril synchronization |
|
4520 |
5554 d 17 h |
jermar |
/trunk/uspace/srv/vfs/ |
Synchronize access to pending_req. |
|
4519 |
5554 d 21 h |
decky |
/trunk/uspace/srv/vfs/ |
remove yet another serialization which is not needed anymore |
|
4518 |
5555 d 13 h |
jermar |
/trunk/uspace/srv/vfs/ |
Make VFS use the new synchronization for fibrils. Now there should be no (or
only secondary) fibril serialization. Code reorganized not to hold the phone
lock during async_wait_for() in most cases. Tested on ia32. On amd64, VFS
crashes, but I think it is an unrelated problem. |
|
4509 |
5556 d 22 h |
decky |
/trunk/uspace/ |
rename libadt -> adt |
|
4507 |
5558 d 13 h |
svoboda |
/trunk/ |
Primitive ATA disk driver that you can use in Qemu (read only). |
|
4504 |
5559 d 23 h |
decky |
/trunk/ |
group fhc and obio under a common directory |
|
4497 |
5560 d 18 h |
decky |
/trunk/uspace/srv/console/ |
no not disturb console initialization by connections |
|
4496 |
5560 d 19 h |
decky |
/trunk/uspace/ |
be more pedantic about undefined function prototypes (to avoid memory access traps)
add missing header files |
|
4494 |
5560 d 23 h |
decky |
/trunk/uspace/srv/console/ |
keep the kernel console intact as long as it is possible (to be able to see any out-of-order errors) |
|
4493 |
5561 d 0 h |
decky |
/trunk/uspace/srv/console/ |
add forgotten answer in case of an error |
|
4492 |
5561 d 2 h |
jermar |
/trunk/uspace/ |
inode_t -> fdi_node_t |
|
4491 |
5561 d 13 h |
decky |
/trunk/uspace/ |
rename libc's fs_node_t to inode_t to avoid type conflict with VFS |
|
4465 |
5561 d 17 h |
decky |
/trunk/uspace/srv/fs/ |
implement VFS_OPEN_NODE and VFS_DEVICE (no device handle is returned)
implement no-op VFS_SYNC and VFS_CLOSE |
|
4464 |
5561 d 17 h |
decky |
/trunk/uspace/srv/fs/devfs/ |
devfs rewrite:
- implement VFS_OPEN_NODE, VFS_DEVICE, VFS_SYNC, VFS_CLOSE
- connect to underlying device during VFS_LOOKUP (if L_OPEN flag is used) and VFS_OPEN_NODE
- disconnect from underlying device during VFS_CLOSE
- forward VFS_READ, VFS_WRITE and VFS_SYNC to the underlying device
- VFS_DEVICE: return underlying device handle |
|
4463 |
5561 d 17 h |
decky |
/trunk/uspace/srv/vfs/ |
VFS rewrite:
- add new VFS methods
- VFS_NODE for getting VFS node for a file descriptor
- VFS_OPEN_NODE for opening an i-node directly
- VFS_DEVICE for getting the underlying device of a file (if any)
- VFS_SYNC for syncing buffers
- L_OPEN flag is set to underlying filesystem in VFS_LOOKUP when opening a file to support initialization (if any)
- VFS_CLOSE is now propagated to underlying filesystem to support cleanup (if any) |
|
4462 |
5561 d 17 h |
decky |
/trunk/uspace/srv/kbd/ |
keyboard server rewrite: rename "kbd events" to more generic "console events" and similar changes
rename key_buffer.* to keybuffer.*
coding style changes |
|
4461 |
5561 d 17 h |
decky |
/trunk/uspace/srv/loader/ |
add support for passing preset files
rename loader_ prefix to ldr_ prefix (to avoid confusion with libc functions) |
|
4459 |
5561 d 17 h |
decky |
/trunk/uspace/srv/ns/ |
split naming service into multiple files for better readability
add support for waiting on tasks (receives kernel task notifications) |
|
4458 |
5561 d 17 h |
decky |
/trunk/uspace/srv/devmap/ |
remove futex (devmap is not multithreaded), use fibril serialization instead |
|
4457 |
5561 d 17 h |
decky |
/trunk/uspace/srv/fb/ |
framebuffer server rewrite: cleanup, always use canonical order of coordinates (x, y) |
|
4456 |
5561 d 17 h |
decky |
/trunk/uspace/srv/console/ |
console server rewrite: use VFS_READ/VFS_WRITE for generic I/O, register separate virtual consoles using device mapper |
|
4445 |
5563 d 19 h |
decky |
/trunk/uspace/srv/vfs/ |
revert commit 4443 and implement a much finer solution as suggested by Jakub Jermar in commit 4444 |
|
4444 |
5566 d 16 h |
jermar |
/trunk/uspace/srv/vfs/ |
Put a comment for consideration. |
|
4443 |
5566 d 18 h |
decky |
/trunk/uspace/srv/vfs/ |
serialization of VFS_OPEN implementation to avoid lockup by concurrent open() operations from multiple tasks |
|
4442 |
5567 d 9 h |
svoboda |
/trunk/uspace/srv/bd/gxe_bd/ |
Support more than one disk in gxe_bd (hardcoded to 2). |
|
4440 |
5567 d 16 h |
jermar |
/trunk/uspace/srv/bd/gxe_bd/ |
Large I/O areas actually work as expected. Fix the definition of the gxe_bd device and pio_enable() the whole I/O area at once. |
|
4437 |
5568 d 2 h |
jermar |
/trunk/uspace/srv/vfs/ |
Fix function definition. |
|
4435 |
5568 d 17 h |
jermar |
/trunk/uspace/ |
More conversions to dev_handle_t. |
|
4432 |
5572 d 19 h |
decky |
/trunk/uspace/srv/vfs/ |
avoid accessing uninitialized files array at all times |
|
4431 |
5572 d 19 h |
jermar |
/trunk/uspace/srv/vfs/ |
Hold the namespace rwlock during the entire operation of vfs_mount_internal().
Get rid of rootfs futex, which is not needed anymore.
Do not instantiate VFS node for the mounted root if the mount failed. |
|