On Tue, Jul 25, 2017 at 06:26:06PM +0700, Nikolay Marchuk wrote: > Hello, strace community! > > Project: Advanced syscall filtering syntax. > > > Accomplishments: this week I've been working on improvements of fd > filtering. Unfortunately, I haven't found good solution for the issue > with negative fds, but I've found and fixed other issues of fd > filtering. In my opinion, the only solution for the issue with negative > fds is to rewrite number_set, because bitsets are not suitable for > storing negative numbers, but I want to collect feedback on this solution. Thinking of this, I'd prefer seeing some B*-tree/trie based approach for storing these masks (looks like only 3 levels needed in order to store 32-bit mask and 6 for 64-bit mask), which then can also be used for filtering *ctl commands in the future, but I think that the negative fd issue is not that critical anyway (as fds are unsigned in kernel and checked against rlimit which can't be greater than INT_MAX, the case of filtering for negative fds is pretty exotic).
------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Strace-devel mailing list Strace-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/strace-devel
