... (Bill Maniatty) >Subject: Re: shared libs - can everyone be happy with this? > ... code space, otherwise how would >|> they be loaded in. > [Stuff Deleted] > >I don't ... 's address space (different Intel segments, with descriptors in the LDT, each ...

linux-activists - Drew Eckhardt - Aug 17 1992 - 23:15

... (Bill Maniatty) Subject: Re: shared libs - can everyone be happy with this? Message- ... code space, otherwise how would |> they be loaded in. [Stuff Deleted] I don ... support? Bill |> Nathan I. Laredo (this spot: $9) Internet: gt7080a@prism. ...

linux-activists - Bill Maniatty - Aug 17 1992 - 14:09

... the usual rules of shared or not memory. This memory doesn't get ... and therefore it is equally valid with or without the mod and can ... to the end of the lib, and then the jump table ... as I don't have to use jump tables I can be happy, although I hope people ...

linux-activists - william E Davidsen - Aug 17 1992 - 14:53

... the latest test release. > >Bravo! Then everyone will be able to make the best ... in the Libs. Since the kernel is still evolving I can understand at ... long as this is valid I have absolutely no concern with Jump tables except occasionly the perfomance ...

linux-activists - Hubert Bahr - Aug 15 1992 - 03:13

... is no longer shared. There is a > term for this that you probably ... using the page would be using the same libs. Of course, some kernel ... t recall seeing anything to prevent this). > David > -- > David Engel Optical Data Systems ...

linux-activists - Mika Pekka Liljeberg - Aug 17 1992 - 14:33

... no longer shared. There is a >> term for this that you ... the > page would be using the same libs. Of course, some ... t change too profoundly. > > Would this break demand loading, then? I ... in ( *BOOM* ). Try to find that one with your debugger! :-) Rick ...

linux-activists - nichols - Aug 21 1992 - 17:24

... understand at least some change in the Libs. However, I understand : that RS is working ... way. So recompiling already working code with the new version won't gain ... that problem too? Nope, jump tables can't do anything about application changes! ...

linux-activists - David Engel - Aug 15 1992 - 13:10

... (David Engel) writes: > No, it wouldn't be shareable. When a process writes to a shared ... problems if a different version of the same shared library is installed and you better make sure that this mechanism can't be used to subvert system security. Might ...

linux-activists - Werner Almesberger - Aug 17 1992 - 14:25

... hard CPU-seconds! I'll be timing again >| this evening, and might even ... have to re-download it), but I retried three different programs with a ... , which should do about 1E5 lib-calls, is convincing to me. I ... 4.17u 5.64s 9.91 (shared) : 4.12u 5.57s 9.90 ...

linux-activists - Rob Hooft - Aug 20 1992 - 06:46

... . So, instead of having > 200K shared between all : | of your shells, ... for all processes it would still be sharable. And it would not ... valid. No, it wouldn't be shareable. When a process writes to ... no longer shared. There is a term for this that you ...

linux-activists - David Engel - Aug 17 1992 - 11:13

... So, instead of having > 200K shared between all : | of your shells ... this was done without guile. A brute force technique would be ... the page. Note that with processes sharing the same text, there ... m not sure that all this trouble is worth it. Does ...

linux-activists - Jay A. Carlson - Aug 17 1992 - 13:50

... give each process its own copy if this was done | without guile. Exactly so. ... segs, they better be clean. | Note that with processes sharing the same text, ... point per process. I had completely missed this. The overhead of a single fixup is ...

linux-activists - william E Davidsen - Aug 18 1992 - 08:59

... colorado.edu (Drew Eckhardt) writes: | With shared pages, COW is implemented ... asking about memory protection, you can find some nice discussions of ... a pretty good one) will be an overview, since a good ... coverage of the topic would be about 40 pages of print ...

linux-activists - william E Davidsen - Aug 18 1992 - 09:04

Is the jump table stuff going back to FSF to become part of main-line GCC? Or is there something linux-specific in the implementation? -- David Boyce dsb@world.std.com 617-576-1540

linux-activists - David Boyce - Aug 15 1992 - 12:36

... So, instead of having > 200K shared between all | of your shells, ... all processes it would still be sharable. And it would not ... kernel does not currently support this is valid, and I don' ... reason why that would not be satisfactory in terms of performance ...

linux-activists - william E Davidsen - Aug 17 1992 - 10:47

... by over 0.5 seconds. We're talking a | program here that runs ... hard CPU-seconds! I'll be timing again | this evening, and might even ... . I guess I'll be using jump-libs for the rest of ... jump tables before I have to be suspicious. If more instructions and ...

linux-activists - william E Davidsen - Aug 19 1992 - 08:55

... assumes data space is executable (it can build small functions on the ... /still/ seems to fail despite being able to try, is a ... that the standard library should be well-behaved. But I'm ... -modifying code, maybe it should be permissible. I don't dislike ...

linux-activists - Linus Benedict Torvalds - Aug 21 1992 - 03:55

... don't load the module : and you get linked to the real routines directly in memory in the slib. : The advantage of this is that you get all the benefits of the current : slib implementation in terms of performance. Actually, we've already done that ...

linux-activists - David Engel - Aug 14 1992 - 12:22

... estdio library |> was broken, and wrote to the code. Does linux permit self modifying code (by design or bug)? This seems like a pretty bad ``feature''. Bill -- | | maniattb@cs.rpi.edu - in real life Bill Maniatty |

linux-activists - Bill Maniatty - Aug 17 1992 - 10:42

... jump tables? >| >| I didn't believe this, so I repeated it many times. ... the crt0.o? > Have you timed this with the time command (I posted ... by over 0.5 seconds. We're talking a program here that runs ... . I guess I'll be using jump-libs for the rest of ...

linux-activists - Rob Hooft - Aug 18 1992 - 11:41