Stephen Kell
Stephen Kell
For anonymous structures, a declaration's line number affects [what we consider as] its name. Its name feeds in to the type summary code, so we can easily get two spuriously...
Linux explicitly sanctions a method of 'setting the process name' which not only clobbers argument and environment strings, but also shifts the effective boundary between the two, such that the...
The concept of [debuginfod](https://sourceware.org/elfutils/Debuginfod.html) is a great fit for liballocs: rather than requiring users to pre-generate the `-meta.so` file for each binary that gets loaded, the debug info could be...
I just noticed that the 'alloca' test case is clearly invalid C, because it declares same-named local variables in the same block but with different types. How can this have...
We're very close to being able to ditch the wrapper script and move to 'toolsub'. The linker plugin(s) should do the following - 'link-used-types' fixup of .o files - globalize...
These patches add support for a serial console. The UART code is loosely based on that of MIT's xv6. It has allowed me at King's to run a coursework that...
The ability to map a loaded DSO `MAP_SHARED` seems powerful as a primitive. It's worth thinking about this more generally. It brings some image-like qualities to Unix systems. Does it...
Currently we leave temporary shared objects lying around the filesystem, which is not good. If they were POSIX shared-memory objects, the cleanup semantics might be different/nicer and a pathname might...
We effectlvely take a snapshot of the kernel's page, which is wrong when the data might change. We need to divert references to our copy so that they point (cross-DSO)...
Currently we assume that the `new_stack` passed to `clone()` is ours to mess with. However, it's conceivable that the `clone()` site has pre-populated the stack with stuff that will be...