Searched hist:50548 (Results 1 - 4 of 4) sorted by relevance

/freebsd-11-stable/sys/dev/sf/
H A Dif_sf.cdiff 50548 Sun Aug 29 07:03:58 MDT 1999 bde Don't restrict our requests for contiguous memory to addresses >= 1MB.
This fixes, at least, panics in ncr_attach() on i386's with about 5MB
of memory. The restriction was a hack to leave some low memory for ISA
DMA, but on i386's we now allocate pages from the top down, so all the
restriction did was cause our allocations to fail when there is no free
memory above 1MB.
/freebsd-11-stable/sys/dev/ti/
H A Dif_ti.cdiff 50548 Sun Aug 29 07:03:58 MDT 1999 bde Don't restrict our requests for contiguous memory to addresses >= 1MB.
This fixes, at least, panics in ncr_attach() on i386's with about 5MB
of memory. The restriction was a hack to leave some low memory for ISA
DMA, but on i386's we now allocate pages from the top down, so all the
restriction did was cause our allocations to fail when there is no free
memory above 1MB.
/freebsd-11-stable/sys/dev/ncr/
H A Dncr.cdiff 50548 Sun Aug 29 07:03:58 MDT 1999 bde Don't restrict our requests for contiguous memory to addresses >= 1MB.
This fixes, at least, panics in ncr_attach() on i386's with about 5MB
of memory. The restriction was a hack to leave some low memory for ISA
DMA, but on i386's we now allocate pages from the top down, so all the
restriction did was cause our allocations to fail when there is no free
memory above 1MB.
/freebsd-11-stable/sys/dev/rl/
H A Dif_rl.cdiff 50548 Sun Aug 29 07:03:58 MDT 1999 bde Don't restrict our requests for contiguous memory to addresses >= 1MB.
This fixes, at least, panics in ncr_attach() on i386's with about 5MB
of memory. The restriction was a hack to leave some low memory for ISA
DMA, but on i386's we now allocate pages from the top down, so all the
restriction did was cause our allocations to fail when there is no free
memory above 1MB.

Completed in 186 milliseconds