diff options
| author | Glenn Morris | 2014-01-06 21:11:14 -0500 |
|---|---|---|
| committer | Glenn Morris | 2014-01-06 21:11:14 -0500 |
| commit | 05834033161a7769b02dd2eea976938bb1795733 (patch) | |
| tree | 341de1e82d8e18617baff82837331fa432528d59 | |
| parent | 0d3c6661753ce5fd9f53ff9306b90a73addb8ef2 (diff) | |
| download | emacs-05834033161a7769b02dd2eea976938bb1795733.tar.gz emacs-05834033161a7769b02dd2eea976938bb1795733.zip | |
* etc/PROBLEMS: More removal of old stuff
| -rw-r--r-- | etc/PROBLEMS | 58 |
1 files changed, 5 insertions, 53 deletions
diff --git a/etc/PROBLEMS b/etc/PROBLEMS index f3e2b5d6689..3d7a426b3fa 100644 --- a/etc/PROBLEMS +++ b/etc/PROBLEMS | |||
| @@ -400,7 +400,7 @@ You need to configure your machine with a fully qualified domain name, | |||
| 400 | If you cannot fix the configuration, you can set the Lisp variable | 400 | If you cannot fix the configuration, you can set the Lisp variable |
| 401 | mail-host-address to the value you want. | 401 | mail-host-address to the value you want. |
| 402 | 402 | ||
| 403 | ** NFS and RFS | 403 | ** NFS |
| 404 | 404 | ||
| 405 | *** Emacs says it has saved a file, but the file does not actually | 405 | *** Emacs says it has saved a file, but the file does not actually |
| 406 | appear on disk. | 406 | appear on disk. |
| @@ -412,55 +412,6 @@ detect the problem. Emacs checks the failure codes of all the system | |||
| 412 | calls involved in writing a file, including `close'; but in the case | 412 | calls involved in writing a file, including `close'; but in the case |
| 413 | where the problem occurs, none of those system calls fails. | 413 | where the problem occurs, none of those system calls fails. |
| 414 | 414 | ||
| 415 | *** Editing files through RFS gives spurious "file has changed" warnings. | ||
| 416 | It is possible that a change in Emacs 18.37 gets around this problem, | ||
| 417 | but in case not, here is a description of how to fix the RFS bug that | ||
| 418 | causes it. | ||
| 419 | |||
| 420 | There was a serious pair of bugs in the handling of the fsync() system | ||
| 421 | call in the RFS server. | ||
| 422 | |||
| 423 | The first is that the fsync() call is handled as another name for the | ||
| 424 | close() system call (!!). It appears that fsync() is not used by very | ||
| 425 | many programs; Emacs version 18 does an fsync() before closing files | ||
| 426 | to make sure that the bits are on the disk. | ||
| 427 | |||
| 428 | This is fixed by the enclosed patch to the RFS server. | ||
| 429 | |||
| 430 | The second, more serious problem, is that fsync() is treated as a | ||
| 431 | non-blocking system call (i.e., it's implemented as a message that | ||
| 432 | gets sent to the remote system without waiting for a reply). Fsync is | ||
| 433 | a useful tool for building atomic file transactions. Implementing it | ||
| 434 | as a non-blocking RPC call (when the local call blocks until the sync | ||
| 435 | is done) is a bad idea; unfortunately, changing it will break the RFS | ||
| 436 | protocol. No fix was supplied for this problem. | ||
| 437 | |||
| 438 | (as always, your line numbers may vary) | ||
| 439 | |||
| 440 | % rcsdiff -c -r1.2 serversyscall.c | ||
| 441 | RCS file: RCS/serversyscall.c,v | ||
| 442 | retrieving revision 1.2 | ||
| 443 | diff -c -r1.2 serversyscall.c | ||
| 444 | *** /tmp/,RCSt1003677 Wed Jan 28 15:15:02 1987 | ||
| 445 | --- serversyscall.c Wed Jan 28 15:14:48 1987 | ||
| 446 | *************** | ||
| 447 | *** 163,169 **** | ||
| 448 | /* | ||
| 449 | * No return sent for close or fsync! | ||
| 450 | */ | ||
| 451 | ! if (syscall == RSYS_close || syscall == RSYS_fsync) | ||
| 452 | proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); | ||
| 453 | else | ||
| 454 | { | ||
| 455 | --- 166,172 ---- | ||
| 456 | /* | ||
| 457 | * No return sent for close or fsync! | ||
| 458 | */ | ||
| 459 | ! if (syscall == RSYS_close) | ||
| 460 | proc->p_returnval = deallocate_fd(proc, msg->m_args[0]); | ||
| 461 | else | ||
| 462 | { | ||
| 463 | |||
| 464 | ** PSGML conflicts with sgml-mode. | 415 | ** PSGML conflicts with sgml-mode. |
| 465 | 416 | ||
| 466 | PSGML package uses the same names of some variables (like keymap) | 417 | PSGML package uses the same names of some variables (like keymap) |
| @@ -495,9 +446,10 @@ This is a known limitation of Emacs with no solution at this time. | |||
| 495 | 446 | ||
| 496 | *** Emacs uses 100% of CPU time | 447 | *** Emacs uses 100% of CPU time |
| 497 | 448 | ||
| 498 | This is a known problem with some versions of the Semantic package. | 449 | This was a known problem with some old versions of the Semantic package. |
| 499 | The solution is to upgrade Semantic to version 2.0pre4 (distributed | 450 | The solution was to upgrade Semantic to version 2.0pre4 (distributed |
| 500 | with CEDET 1.0pre4) or later. | 451 | with CEDET 1.0pre4) or later. Note that Emacs includes Semantic since |
| 452 | 23.2, and this issue does not apply to the included version. | ||
| 501 | 453 | ||
| 502 | *** Self-documentation messages are garbled. | 454 | *** Self-documentation messages are garbled. |
| 503 | 455 | ||