/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright 2008 Sun Microsystems, Inc. All rights reserved.
* Use is subject to license terms.
*/
#ifndef _NPI_RX_RD32_H
#define _NPI_RX_RD32_H
#pragma ident "%Z%%M% %I% %E% SMI"
#ifdef __cplusplus
extern "C" {
#endif
#include <npi.h>
#pragma inline(RXDMA_REG_READ32)
/*
* RXDMA_REG_READ32
*
* Read a 32-bit value from a DMC register.
*
* Arguments:
* handle The NPI handle to use.
* offset The offset into the DMA CSR (the register).
* channel The channel, which is used as a multiplicand.
*
* Notes:
* If handle.regp is a virtual address (the address of a VR),
* we have to subtract the value DMC right off the bat. DMC
* is defined as 0x600000, which works in a non-virtual address
* space, but not in a VR. In a VR, a DMA CSR's space begins
* at zero (0). So, since every call to RXMDA_REG_READ32 uses
* a register macro which adds in DMC, we have to subtract it.
*
* The rest of it is pretty straighforward. In a VR, a channel is
* logical, not absolute; and every DMA CSR is 512 bytes big;
* furthermore, a subpage of a VR is always ordered with the
* transmit CSRs first, followed by the receive CSRs. That is,
* a 512 byte space of Tx CSRs, followed by a 512 byte space of
* Rx CSRs. Hence this calculation:
*
* offset += ((channel << 1) + 1) << DMA_CSR_SLL;
*
* Here's an example:
*
* RXDMA_REG_READ32(handle, RX_DMA_CTL_STAT_REG, channel);
* Let's say channel is 3
* #define RX_DMA_CTL_STAT_REG (DMC + 0x00070)
* offset = 0x600070
* offset &= 0xff = 0x70
* offset += ((3 << 1) + 1) << 9
* 3 << 1 = 6
* 6 + 1 = 7
* 7 << 9 = 0xe00
* offset += 0xe00 = 0xe70
*
* Therefore, our register's (virtual) PIO address is 0xe70.
*
* cf. Table 10-6 on page 181 of the Neptune PRM, v 1.4:
*
* E00 - FFF CSRs for bound logical receive DMA channel 3.
*
* In a non-virtual environment, you simply multiply the absolute
* channel number by 512 bytes, and get the correct offset to
* the register you're looking for. That is, the RX_DMA_CTL_STAT CSR,
* is, as are all of these registers, in a table where each channel
* is offset 512 bytes from the previous channel (count 16 step 512).
*
* offset += (channel << DMA_CSR_SLL); // channel<<9 = channel*512
*
* Here's an example:
*
* RXDMA_REG_READ32(handle, RX_DMA_CTL_STAT_REG, channel);
* Let's say channel is 3
* #define RX_DMA_CTL_STAT_REG (DMC + 0x00070)
* offset = 0x600070
* offset += (3 << 9)
* 3 << 9 = 0x600
* offset += 0x600 = 0x600670
*
* Therefore, our register's PIO address is 0x600670.
*
* cf. Table 12-42 on page 234 of the Neptune PRM, v 1.4:
* RX_DMA_CTL_STAT (DMC + [0x]00070) (count 16 step [0x]200)
*
* Context:
* Guest domain
*
*/
int channel)
{
offset &= DMA_CSR_MASK;
} else {
}
}
#ifdef __cplusplus
}
#endif
#endif /* _NPI_RX_RD32_H */