On Fri, 2023-11-10 at 11:12 +0000, Anton Ivanov wrote: > > > On 10/11/2023 11:10, Benjamin Berg wrote: > > Hi, > > > > On Fri, 2023-11-10 at 10:56 +0000, Anton Ivanov wrote: > > > On 10/11/2023 10:42, Anton Ivanov wrote: > > > > [SNIP] > > > > > > > > If we are going to use this definition of CATCH_EINTR > > > > throughout we > > > > might as well remove the partial read/write code in UBD and > > > > other > > > > places. > > > > > > Actually - some of it. In some places it will still be needed > > > (trying > > > to get my head around where this will work and where not). > > > > Do you mean that we could just use the os_* helpers here and define > > them to catch EINTR and also deal with partial read()/write()? > > > > Yep :)
Looking at the os_write_file/os_read_file users, the only place where a full read/write does not seem appropriate to me is in hostaudio_kern.c. There are quite a few places where a read/write cannot (or should not) be short, but a retry loop there is completely fine. Benjamin > > > > Benjamin > > > > _______________________________________________ > > linux-um mailing list > > linux-um@lists.infradead.org > > http://lists.infradead.org/mailman/listinfo/linux-um > > > _______________________________________________ linux-um mailing list linux-um@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-um
