3 ms·

Author here. Yeah, there are reasons to use dynamic linking, but I'm still not sure why dynamic linking gives you better profiling and tracing. > Unfortunatel

Author here. Yeah, there are reasons to use dynamic linking, but I'm still not sure why dynamic linking gives you better profiling and tracing. > Unfortunately people typically don't follow Drepper's advice (as far I remember the article) on building shared libraries, particularly using symbol versioning a la glibc. I mention this in the post, but I quoted another person, Ian Lance Taylor, who wrote the gold linker, as saying that symbol combining, a portion of symbol versioning, is an unsolved problem. Even if developers don't follow Drepper's advice, I wouldn't put the blame on them because of that fact; I would argue that it's more accurate to say they can't really follow his advice.

dchapp 5y ago
> Yeah, there are reasons to use dynamic linking, but I'm still not sure why dynamic linking gives you better profiling and tracing. It's not so much that the tracing becomes better, but that it becomes feasible at all. Two specific situations come to mind, both MPI-adjacent. (1) Running a PMPI-based tool on code you can't recompile yourself (e.g., you need a Q clearance to see the source, but not to actually execute it--weird I know, but not that uncommon in the DOE labs.); and (2) running multiple PMPI-based tools simultaneously which are composed at runtime via PnMPI.
- ghoward 5y ago
  I addressed these arguments. I think that if you don't have access to the source, it's a security problem. That includes not having access to the source because of security clearances. As far as I am concerned, your superiors in the DOE are adversaries to you and your machine. Other people will have different viewpoints on that, and that's fine.
  - gnufx 5y ago
    I want free software, and I'm aghast at what some people run, but this is not the real world. How many examples do you want? If you talk security, what's the threat model? What I run on a decently managed (with the aid of dynamically linked libraries) compute cluster should only put my data at risk. As I understood it, there actually has been a push for free software solutions on CORAL systems, but I don't remember where that came from.
- gnufx 5y ago
  Exactly, but even if you can rebuild it, you don't want to under most circumstances. In principle with static binaries you can use dyninst, for instance, but in practice you may not be able to for various reasons. Then as a system manager you'd like to have global profiling of what runs -- for various reasons, including putting the result for a program in front of a user. If it's all dynamically linked, you can hook in and do that by default, and with various levels of insistence, depending on the hooks.
gnufx 5y ago
Drepper's scheme obviously works tolerably well with glibc, in particular, even if he didn't know what he was talking about. If you don't have symbol versioning, and you want to statically link something with parts that depend on incompatible versions of a library, you're obviously stuffed.

dchapp 5y ago

> Yeah, there are reasons to use dynamic linking, but I'm still not sure why dynamic linking gives you better profiling and tracing. It's not so much that the tracing becomes better, but that it becomes feasible at all. Two specific situations come to mind, both MPI-adjacent. (1) Running a PMPI-based tool on code you can't recompile yourself (e.g., you need a Q clearance to see the source, but not to actually execute it--weird I know, but not that uncommon in the DOE labs.); and (2) running multiple PMPI-based tools simultaneously which are composed at runtime via PnMPI.

ghoward 5y ago
I addressed these arguments. I think that if you don't have access to the source, it's a security problem. That includes not having access to the source because of security clearances. As far as I am concerned, your superiors in the DOE are adversaries to you and your machine. Other people will have different viewpoints on that, and that's fine.
- gnufx 5y ago
  I want free software, and I'm aghast at what some people run, but this is not the real world. How many examples do you want? If you talk security, what's the threat model? What I run on a decently managed (with the aid of dynamically linked libraries) compute cluster should only put my data at risk. As I understood it, there actually has been a push for free software solutions on CORAL systems, but I don't remember where that came from.
gnufx 5y ago
Exactly, but even if you can rebuild it, you don't want to under most circumstances. In principle with static binaries you can use dyninst, for instance, but in practice you may not be able to for various reasons. Then as a system manager you'd like to have global profiling of what runs -- for various reasons, including putting the result for a program in front of a user. If it's all dynamically linked, you can hook in and do that by default, and with various levels of insistence, depending on the hooks.

gnufx 5y ago

Drepper's scheme obviously works tolerably well with glibc, in particular, even if he didn't know what he was talking about. If you don't have symbol versioning, and you want to statically link something with parts that depend on incompatible versions of a library, you're obviously stuffed.