Author: cem
Date: Fri Mar 16 18:50:26 2018
New Revision: 331070
URL: https://svnweb.freebsd.org/changeset/base/331070

Log:
  random(4): Poll for signals during large reads
  
  Occasionally poll for signals during large reads of the /dev/u?random
  devices.  This allows cancellation via SIGINT of accidental invocations of
  very large reads.  (A 2GB /dev/random read, which takes about 10 seconds on
  my 2017 AMD Zen processor, can be aborted.)
  
  I believe this behavior was intended since 2014 (r273997), just not fully
  implemented.
  
  This is motivated by a potential getrandom(2) interface that may not
  explicitly forbid extremely large reads on 64-bit platforms -- even larger
  than the 2GB limit imposed on devfs I/O by default.  Such reads, if they are
  to be allowed, should be cancellable by the user or administrator.
  
  Reviewed by:  delphij
  Approved by:  secteam (delphij)
  Sponsored by: Dell EMC Isilon
  Differential Revision:        https://reviews.freebsd.org/D14684

Modified:
  head/sys/dev/random/randomdev.c

Modified: head/sys/dev/random/randomdev.c
==============================================================================
--- head/sys/dev/random/randomdev.c     Fri Mar 16 18:16:31 2018        
(r331069)
+++ head/sys/dev/random/randomdev.c     Fri Mar 16 18:50:26 2018        
(r331070)
@@ -130,7 +130,13 @@ READ_RANDOM_UIO(struct uio *uio, bool nonblock)
        uint8_t *random_buf;
        int error, spamcount;
        ssize_t read_len, total_read, c;
+       /* 16 MiB takes about 0.08 s CPU time on my 2017 AMD Zen CPU */
+#define SIGCHK_PERIOD (16 * 1024 * 1024)
+       const size_t sigchk_period = SIGCHK_PERIOD;
 
+       CTASSERT(SIGCHK_PERIOD % PAGE_SIZE == 0);
+#undef SIGCHK_PERIOD
+
        random_buf = malloc(PAGE_SIZE, M_ENTROPY, M_WAITOK);
        p_random_alg_context->ra_pre_read();
        error = 0;
@@ -167,11 +173,22 @@ READ_RANDOM_UIO(struct uio *uio, bool nonblock)
                        read_len = MIN(read_len, PAGE_SIZE);
                        p_random_alg_context->ra_read(random_buf, read_len);
                        c = MIN(uio->uio_resid, read_len);
+                       /*
+                        * uiomove() may yield the CPU before each 'c' bytes
+                        * (up to PAGE_SIZE) are copied out.
+                        */
                        error = uiomove(random_buf, c, uio);
                        total_read += c;
+                       /*
+                        * Poll for signals every few MBs to avoid very long
+                        * uninterruptible syscalls.
+                        */
+                       if (error == 0 && uio->uio_resid != 0 &&
+                           total_read % sigchk_period == 0)
+                               error = tsleep_sbt(&random_alg_context, PCATCH,
+                                   "randrd", SBT_1NS, 0, C_HARDCLOCK);
                }
-               if (total_read != uio->uio_resid && (error == ERESTART || error 
== EINTR))
-                       /* Return partial read, not error. */
+               if (error == ERESTART || error == EINTR)
                        error = 0;
        }
        free(random_buf, M_ENTROPY);
_______________________________________________
svn-src-head@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-head
To unsubscribe, send any mail to "svn-src-head-unsubscr...@freebsd.org"

Reply via email to