On Sun, 14 Jun 2020, Dorit Mari via curl-library wrote:
I use libcurl version 7.65.3. The OS is Linux. I always use CURLOPT_RESOLVE
...
libcurl compiled without this flag (45% vs. 12%).
Wow, that's seem super strange. How on earth can the resolving code - that you
don't even use - affect the CPU load of the transfer phase? And to that
serious degree!
Does the resolver socket linger around during the transfer for you so that you
monitor and "act" on it too, in addition to the transfer socket?
In which transfer states do you see this increased CPU load, do you know?
Can you show us a small example source code that reproduces the load issue for
you?
Is there any such known problem/solution?
Not that I can think of. It is of course always clever to build with the
latest version as a test shot and see if it makes anything different, it
*might* save us from debugging and chasing ghosts.
--
/ daniel.haxx.se | Commercial curl support up to 24x7 is available!
| Private help, bug fixes, support, ports, new features
| https://www.wolfssl.com/contact/
-------------------------------------------------------------------
Unsubscribe: https://cool.haxx.se/list/listinfo/curl-library
Etiquette: https://curl.haxx.se/mail/etiquette.html
