qemu-e2k/hw/intc/arm_gicv3_its.c

1334 lines
40 KiB
C
Raw Normal View History

/*
* ITS emulation for a GICv3-based system
*
* Copyright Linaro.org 2021
*
* Authors:
* Shashi Mallela <shashi.mallela@linaro.org>
*
* This work is licensed under the terms of the GNU GPL, version 2 or (at your
* option) any later version. See the COPYING file in the top-level directory.
*
*/
#include "qemu/osdep.h"
#include "qemu/log.h"
#include "hw/qdev-properties.h"
#include "hw/intc/arm_gicv3_its_common.h"
#include "gicv3_internal.h"
#include "qom/object.h"
#include "qapi/error.h"
typedef struct GICv3ITSClass GICv3ITSClass;
/* This is reusing the GICv3ITSState typedef from ARM_GICV3_ITS_COMMON */
DECLARE_OBJ_CHECKERS(GICv3ITSState, GICv3ITSClass,
ARM_GICV3_ITS, TYPE_ARM_GICV3_ITS)
struct GICv3ITSClass {
GICv3ITSCommonClass parent_class;
void (*parent_reset)(DeviceState *dev);
};
/*
* This is an internal enum used to distinguish between LPI triggered
* via command queue and LPI triggered via gits_translater write.
*/
typedef enum ItsCmdType {
NONE = 0, /* internal indication for GITS_TRANSLATER write */
CLEAR = 1,
DISCARD = 2,
INTERRUPT = 3,
} ItsCmdType;
typedef struct {
uint32_t iteh;
uint64_t itel;
} IteEntry;
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
/*
* The ITS spec permits a range of CONSTRAINED UNPREDICTABLE options
* if a command parameter is not correct. These include both "stall
* processing of the command queue" and "ignore this command, and
* keep processing the queue". In our implementation we choose that
* memory transaction errors reading the command packet provoke a
* stall, but errors in parameters cause us to ignore the command
* and continue processing.
* The process_* functions which handle individual ITS commands all
* return an ItsCmdResult which tells process_cmdq() whether it should
* stall or keep going.
*/
typedef enum ItsCmdResult {
CMD_STALL = 0,
CMD_CONTINUE = 1,
} ItsCmdResult;
static uint64_t baser_base_addr(uint64_t value, uint32_t page_sz)
{
uint64_t result = 0;
switch (page_sz) {
case GITS_PAGE_SIZE_4K:
case GITS_PAGE_SIZE_16K:
result = FIELD_EX64(value, GITS_BASER, PHYADDR) << 12;
break;
case GITS_PAGE_SIZE_64K:
result = FIELD_EX64(value, GITS_BASER, PHYADDRL_64K) << 16;
result |= FIELD_EX64(value, GITS_BASER, PHYADDRH_64K) << 48;
break;
default:
break;
}
return result;
}
static bool get_cte(GICv3ITSState *s, uint16_t icid, uint64_t *cte,
MemTxResult *res)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t l2t_addr;
uint64_t value;
bool valid_l2t;
uint32_t l2t_id;
uint32_t num_l2_entries;
if (s->ct.indirect) {
l2t_id = icid / (s->ct.page_sz / L1TABLE_ENTRY_SIZE);
value = address_space_ldq_le(as,
s->ct.base_addr +
(l2t_id * L1TABLE_ENTRY_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
if (*res == MEMTX_OK) {
valid_l2t = (value & L2_TABLE_VALID_MASK) != 0;
if (valid_l2t) {
num_l2_entries = s->ct.page_sz / s->ct.entry_sz;
l2t_addr = value & ((1ULL << 51) - 1);
*cte = address_space_ldq_le(as, l2t_addr +
((icid % num_l2_entries) * GITS_CTE_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
}
}
} else {
/* Flat level table */
*cte = address_space_ldq_le(as, s->ct.base_addr +
(icid * GITS_CTE_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
}
return FIELD_EX64(*cte, CTE, VALID);
}
static bool update_ite(GICv3ITSState *s, uint32_t eventid, uint64_t dte,
IteEntry ite)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t itt_addr;
MemTxResult res = MEMTX_OK;
itt_addr = FIELD_EX64(dte, DTE, ITTADDR);
itt_addr <<= ITTADDR_SHIFT; /* 256 byte aligned */
address_space_stq_le(as, itt_addr + (eventid * (sizeof(uint64_t) +
sizeof(uint32_t))), ite.itel, MEMTXATTRS_UNSPECIFIED,
&res);
if (res == MEMTX_OK) {
address_space_stl_le(as, itt_addr + (eventid * (sizeof(uint64_t) +
sizeof(uint32_t))) + sizeof(uint32_t), ite.iteh,
MEMTXATTRS_UNSPECIFIED, &res);
}
if (res != MEMTX_OK) {
return false;
} else {
return true;
}
}
static bool get_ite(GICv3ITSState *s, uint32_t eventid, uint64_t dte,
uint16_t *icid, uint32_t *pIntid, MemTxResult *res)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t itt_addr;
bool status = false;
IteEntry ite = {};
itt_addr = FIELD_EX64(dte, DTE, ITTADDR);
itt_addr <<= ITTADDR_SHIFT; /* 256 byte aligned */
ite.itel = address_space_ldq_le(as, itt_addr +
(eventid * (sizeof(uint64_t) +
sizeof(uint32_t))), MEMTXATTRS_UNSPECIFIED,
res);
if (*res == MEMTX_OK) {
ite.iteh = address_space_ldl_le(as, itt_addr +
(eventid * (sizeof(uint64_t) +
sizeof(uint32_t))) + sizeof(uint32_t),
MEMTXATTRS_UNSPECIFIED, res);
if (*res == MEMTX_OK) {
hw/intc/arm_gicv3_its: Don't misuse GITS_TYPE_PHYSICAL define The GITS_TYPE_PHYSICAL define is the value we set the GITS_TYPER.Physical field to -- this is 1 to indicate that we support physical LPIs. (Support for virtual LPIs is the GITS_TYPER.Virtual field.) We also use this define as the *value* that we write into an interrupt translation table entry's INTTYPE field, which should be 1 for a physical interrupt and 0 for a virtual interrupt. Finally, we use it as a *mask* when we read the interrupt translation table entry INTTYPE field. Untangle this confusion: define an ITE_INTTYPE_VIRTUAL and ITE_INTTYPE_PHYSICAL to be the valid values of the ITE INTTYPE field, and replace the ad-hoc collection of ITE_ENTRY_* defines with use of the FIELD() macro to define the fields of an ITE and the FIELD_EX64() and FIELD_DP64() macros to read and write them. We use ITE in the new setup, rather than ITE_ENTRY, because ITE stands for "Interrupt translation entry" and so the extra "entry" would be redundant. We take the opportunity to correct the name of the field that holds the GICv4 'doorbell' interrupt ID (this is always the value 1023 in a GICv3, which is why we were calling it the 'spurious' field). The GITS_TYPE_PHYSICAL define is then used in only one place, where we set the initial GITS_TYPER value. Since GITS_TYPER.Physical is essentially a boolean, hiding the '1' value behind a macro is more confusing than helpful, so expand out the macro there and remove the define entirely. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2022-01-07 18:07:59 +01:00
if (FIELD_EX64(ite.itel, ITE_L, VALID)) {
int inttype = FIELD_EX64(ite.itel, ITE_L, INTTYPE);
if (inttype == ITE_INTTYPE_PHYSICAL) {
*pIntid = FIELD_EX64(ite.itel, ITE_L, INTID);
*icid = FIELD_EX32(ite.iteh, ITE_H, ICID);
status = true;
}
}
}
}
return status;
}
static uint64_t get_dte(GICv3ITSState *s, uint32_t devid, MemTxResult *res)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t l2t_addr;
uint64_t value;
bool valid_l2t;
uint32_t l2t_id;
uint32_t num_l2_entries;
if (s->dt.indirect) {
l2t_id = devid / (s->dt.page_sz / L1TABLE_ENTRY_SIZE);
value = address_space_ldq_le(as,
s->dt.base_addr +
(l2t_id * L1TABLE_ENTRY_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
if (*res == MEMTX_OK) {
valid_l2t = (value & L2_TABLE_VALID_MASK) != 0;
if (valid_l2t) {
num_l2_entries = s->dt.page_sz / s->dt.entry_sz;
l2t_addr = value & ((1ULL << 51) - 1);
value = address_space_ldq_le(as, l2t_addr +
((devid % num_l2_entries) * GITS_DTE_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
}
}
} else {
/* Flat level table */
value = address_space_ldq_le(as, s->dt.base_addr +
(devid * GITS_DTE_SIZE),
MEMTXATTRS_UNSPECIFIED, res);
}
return value;
}
/*
* This function handles the processing of following commands based on
* the ItsCmdType parameter passed:-
* 1. triggering of lpi interrupt translation via ITS INT command
* 2. triggering of lpi interrupt translation via gits_translater register
* 3. handling of ITS CLEAR command
* 4. handling of ITS DISCARD command
*/
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
static ItsCmdResult process_its_cmd(GICv3ITSState *s, uint64_t value,
uint32_t offset, ItsCmdType cmd)
{
AddressSpace *as = &s->gicv3->dma_as;
uint32_t devid, eventid;
MemTxResult res = MEMTX_OK;
bool dte_valid;
uint64_t dte = 0;
uint64_t num_eventids;
uint16_t icid = 0;
uint32_t pIntid = 0;
bool ite_valid = false;
uint64_t cte = 0;
bool cte_valid = false;
uint64_t rdbase;
if (cmd == NONE) {
devid = offset;
} else {
devid = ((value & DEVID_MASK) >> DEVID_SHIFT);
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
}
if (res != MEMTX_OK) {
return CMD_STALL;
}
eventid = (value & EVENTID_MASK);
dte = get_dte(s, devid, &res);
if (res != MEMTX_OK) {
return CMD_STALL;
}
dte_valid = FIELD_EX64(dte, DTE, VALID);
if (!dte_valid) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid command attributes: "
"invalid dte: %"PRIx64" for %d\n",
__func__, dte, devid);
return CMD_CONTINUE;
}
num_eventids = 1ULL << (FIELD_EX64(dte, DTE, SIZE) + 1);
ite_valid = get_ite(s, eventid, dte, &icid, &pIntid, &res);
if (res != MEMTX_OK) {
return CMD_STALL;
}
if (!ite_valid) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid command attributes: invalid ITE\n",
__func__);
return CMD_CONTINUE;
}
cte_valid = get_cte(s, icid, &cte, &res);
if (res != MEMTX_OK) {
return CMD_STALL;
}
if (!cte_valid) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid command attributes: "
"invalid cte: %"PRIx64"\n",
__func__, cte);
return CMD_CONTINUE;
}
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if (devid >= s->dt.num_ids) {
qemu_log_mask(LOG_GUEST_ERROR,
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
"%s: invalid command attributes: devid %d>=%d",
__func__, devid, s->dt.num_ids);
return CMD_CONTINUE;
}
if (eventid >= num_eventids) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid command attributes: eventid %d >= %"
PRId64 "\n",
__func__, eventid, num_eventids);
return CMD_CONTINUE;
}
/*
* Current implementation only supports rdbase == procnum
* Hence rdbase physical address is ignored
*/
rdbase = FIELD_EX64(cte, CTE, RDBASE);
if (rdbase >= s->gicv3->num_cpu) {
return CMD_CONTINUE;
}
if ((cmd == CLEAR) || (cmd == DISCARD)) {
gicv3_redist_process_lpi(&s->gicv3->cpu[rdbase], pIntid, 0);
} else {
gicv3_redist_process_lpi(&s->gicv3->cpu[rdbase], pIntid, 1);
}
if (cmd == DISCARD) {
IteEntry ite = {};
/* remove mapping from interrupt translation table */
return update_ite(s, eventid, dte, ite) ? CMD_CONTINUE : CMD_STALL;
}
return CMD_CONTINUE;
}
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
static ItsCmdResult process_mapti(GICv3ITSState *s, uint64_t value,
uint32_t offset, bool ignore_pInt)
{
AddressSpace *as = &s->gicv3->dma_as;
uint32_t devid, eventid;
uint32_t pIntid = 0;
uint64_t num_eventids;
uint32_t num_intids;
bool dte_valid;
MemTxResult res = MEMTX_OK;
uint16_t icid = 0;
uint64_t dte = 0;
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
ItsCmdResult result = CMD_STALL;
devid = ((value & DEVID_MASK) >> DEVID_SHIFT);
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return result;
}
eventid = (value & EVENTID_MASK);
if (ignore_pInt) {
pIntid = eventid;
} else {
pIntid = ((value & pINTID_MASK) >> pINTID_SHIFT);
}
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return result;
}
icid = value & ICID_MASK;
dte = get_dte(s, devid, &res);
if (res != MEMTX_OK) {
return result;
}
dte_valid = FIELD_EX64(dte, DTE, VALID);
num_eventids = 1ULL << (FIELD_EX64(dte, DTE, SIZE) + 1);
num_intids = 1ULL << (GICD_TYPER_IDBITS + 1);
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if ((devid >= s->dt.num_ids) || (icid >= s->ct.num_ids)
|| !dte_valid || (eventid >= num_eventids) ||
(((pIntid < GICV3_LPI_INTID_START) || (pIntid >= num_intids)) &&
(pIntid != INTID_SPURIOUS))) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid command attributes "
"devid %d or icid %d or eventid %d or pIntid %d or"
"unmapped dte %d\n", __func__, devid, icid, eventid,
pIntid, dte_valid);
/*
* in this implementation, in case of error
* we ignore this command and move onto the next
* command in the queue
*/
} else {
/* add ite entry to interrupt translation table */
hw/intc/arm_gicv3_its: Don't misuse GITS_TYPE_PHYSICAL define The GITS_TYPE_PHYSICAL define is the value we set the GITS_TYPER.Physical field to -- this is 1 to indicate that we support physical LPIs. (Support for virtual LPIs is the GITS_TYPER.Virtual field.) We also use this define as the *value* that we write into an interrupt translation table entry's INTTYPE field, which should be 1 for a physical interrupt and 0 for a virtual interrupt. Finally, we use it as a *mask* when we read the interrupt translation table entry INTTYPE field. Untangle this confusion: define an ITE_INTTYPE_VIRTUAL and ITE_INTTYPE_PHYSICAL to be the valid values of the ITE INTTYPE field, and replace the ad-hoc collection of ITE_ENTRY_* defines with use of the FIELD() macro to define the fields of an ITE and the FIELD_EX64() and FIELD_DP64() macros to read and write them. We use ITE in the new setup, rather than ITE_ENTRY, because ITE stands for "Interrupt translation entry" and so the extra "entry" would be redundant. We take the opportunity to correct the name of the field that holds the GICv4 'doorbell' interrupt ID (this is always the value 1023 in a GICv3, which is why we were calling it the 'spurious' field). The GITS_TYPE_PHYSICAL define is then used in only one place, where we set the initial GITS_TYPER value. Since GITS_TYPER.Physical is essentially a boolean, hiding the '1' value behind a macro is more confusing than helpful, so expand out the macro there and remove the define entirely. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2022-01-07 18:07:59 +01:00
IteEntry ite = {};
ite.itel = FIELD_DP64(ite.itel, ITE_L, VALID, dte_valid);
ite.itel = FIELD_DP64(ite.itel, ITE_L, INTTYPE, ITE_INTTYPE_PHYSICAL);
ite.itel = FIELD_DP64(ite.itel, ITE_L, INTID, pIntid);
hw/intc/arm_gicv3_its: Don't misuse GITS_TYPE_PHYSICAL define The GITS_TYPE_PHYSICAL define is the value we set the GITS_TYPER.Physical field to -- this is 1 to indicate that we support physical LPIs. (Support for virtual LPIs is the GITS_TYPER.Virtual field.) We also use this define as the *value* that we write into an interrupt translation table entry's INTTYPE field, which should be 1 for a physical interrupt and 0 for a virtual interrupt. Finally, we use it as a *mask* when we read the interrupt translation table entry INTTYPE field. Untangle this confusion: define an ITE_INTTYPE_VIRTUAL and ITE_INTTYPE_PHYSICAL to be the valid values of the ITE INTTYPE field, and replace the ad-hoc collection of ITE_ENTRY_* defines with use of the FIELD() macro to define the fields of an ITE and the FIELD_EX64() and FIELD_DP64() macros to read and write them. We use ITE in the new setup, rather than ITE_ENTRY, because ITE stands for "Interrupt translation entry" and so the extra "entry" would be redundant. We take the opportunity to correct the name of the field that holds the GICv4 'doorbell' interrupt ID (this is always the value 1023 in a GICv3, which is why we were calling it the 'spurious' field). The GITS_TYPE_PHYSICAL define is then used in only one place, where we set the initial GITS_TYPER value. Since GITS_TYPER.Physical is essentially a boolean, hiding the '1' value behind a macro is more confusing than helpful, so expand out the macro there and remove the define entirely. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2022-01-07 18:07:59 +01:00
ite.itel = FIELD_DP64(ite.itel, ITE_L, DOORBELL, INTID_SPURIOUS);
ite.iteh = FIELD_DP32(ite.iteh, ITE_H, ICID, icid);
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
result = update_ite(s, eventid, dte, ite) ? CMD_CONTINUE : CMD_STALL;
}
return result;
}
static bool update_cte(GICv3ITSState *s, uint16_t icid, bool valid,
uint64_t rdbase)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t value;
uint64_t l2t_addr;
bool valid_l2t;
uint32_t l2t_id;
uint32_t num_l2_entries;
uint64_t cte = 0;
MemTxResult res = MEMTX_OK;
if (!s->ct.valid) {
return true;
}
if (valid) {
/* add mapping entry to collection table */
cte = FIELD_DP64(cte, CTE, VALID, 1);
cte = FIELD_DP64(cte, CTE, RDBASE, rdbase);
}
/*
* The specification defines the format of level 1 entries of a
* 2-level table, but the format of level 2 entries and the format
* of flat-mapped tables is IMPDEF.
*/
if (s->ct.indirect) {
l2t_id = icid / (s->ct.page_sz / L1TABLE_ENTRY_SIZE);
value = address_space_ldq_le(as,
s->ct.base_addr +
(l2t_id * L1TABLE_ENTRY_SIZE),
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return false;
}
valid_l2t = (value & L2_TABLE_VALID_MASK) != 0;
if (valid_l2t) {
num_l2_entries = s->ct.page_sz / s->ct.entry_sz;
l2t_addr = value & ((1ULL << 51) - 1);
address_space_stq_le(as, l2t_addr +
((icid % num_l2_entries) * GITS_CTE_SIZE),
cte, MEMTXATTRS_UNSPECIFIED, &res);
}
} else {
/* Flat level table */
address_space_stq_le(as, s->ct.base_addr + (icid * GITS_CTE_SIZE),
cte, MEMTXATTRS_UNSPECIFIED, &res);
}
if (res != MEMTX_OK) {
return false;
} else {
return true;
}
}
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
static ItsCmdResult process_mapc(GICv3ITSState *s, uint32_t offset)
{
AddressSpace *as = &s->gicv3->dma_as;
uint16_t icid;
uint64_t rdbase;
bool valid;
MemTxResult res = MEMTX_OK;
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
ItsCmdResult result = CMD_STALL;
uint64_t value;
offset += NUM_BYTES_IN_DW;
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return result;
}
icid = value & ICID_MASK;
rdbase = (value & R_MAPC_RDBASE_MASK) >> R_MAPC_RDBASE_SHIFT;
rdbase &= RDBASE_PROCNUM_MASK;
valid = (value & CMD_FIELD_VALID_MASK);
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if ((icid >= s->ct.num_ids) || (rdbase >= s->gicv3->num_cpu)) {
qemu_log_mask(LOG_GUEST_ERROR,
"ITS MAPC: invalid collection table attributes "
"icid %d rdbase %" PRIu64 "\n", icid, rdbase);
/*
* in this implementation, in case of error
* we ignore this command and move onto the next
* command in the queue
*/
} else {
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
result = update_cte(s, icid, valid, rdbase) ? CMD_CONTINUE : CMD_STALL;
}
return result;
}
static bool update_dte(GICv3ITSState *s, uint32_t devid, bool valid,
uint8_t size, uint64_t itt_addr)
{
AddressSpace *as = &s->gicv3->dma_as;
uint64_t value;
uint64_t l2t_addr;
bool valid_l2t;
uint32_t l2t_id;
uint32_t num_l2_entries;
uint64_t dte = 0;
MemTxResult res = MEMTX_OK;
if (s->dt.valid) {
if (valid) {
/* add mapping entry to device table */
dte = FIELD_DP64(dte, DTE, VALID, 1);
dte = FIELD_DP64(dte, DTE, SIZE, size);
dte = FIELD_DP64(dte, DTE, ITTADDR, itt_addr);
}
} else {
return true;
}
/*
* The specification defines the format of level 1 entries of a
* 2-level table, but the format of level 2 entries and the format
* of flat-mapped tables is IMPDEF.
*/
if (s->dt.indirect) {
l2t_id = devid / (s->dt.page_sz / L1TABLE_ENTRY_SIZE);
value = address_space_ldq_le(as,
s->dt.base_addr +
(l2t_id * L1TABLE_ENTRY_SIZE),
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return false;
}
valid_l2t = (value & L2_TABLE_VALID_MASK) != 0;
if (valid_l2t) {
num_l2_entries = s->dt.page_sz / s->dt.entry_sz;
l2t_addr = value & ((1ULL << 51) - 1);
address_space_stq_le(as, l2t_addr +
((devid % num_l2_entries) * GITS_DTE_SIZE),
dte, MEMTXATTRS_UNSPECIFIED, &res);
}
} else {
/* Flat level table */
address_space_stq_le(as, s->dt.base_addr + (devid * GITS_DTE_SIZE),
dte, MEMTXATTRS_UNSPECIFIED, &res);
}
if (res != MEMTX_OK) {
return false;
} else {
return true;
}
}
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
static ItsCmdResult process_mapd(GICv3ITSState *s, uint64_t value,
uint32_t offset)
{
AddressSpace *as = &s->gicv3->dma_as;
uint32_t devid;
uint8_t size;
uint64_t itt_addr;
bool valid;
MemTxResult res = MEMTX_OK;
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
ItsCmdResult result = CMD_STALL;
devid = ((value & DEVID_MASK) >> DEVID_SHIFT);
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return result;
}
size = (value & SIZE_MASK);
offset += NUM_BYTES_IN_DW;
value = address_space_ldq_le(as, s->cq.base_addr + offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
return result;
}
itt_addr = (value & ITTADDR_MASK) >> ITTADDR_SHIFT;
valid = (value & CMD_FIELD_VALID_MASK);
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if ((devid >= s->dt.num_ids) ||
(size > FIELD_EX64(s->typer, GITS_TYPER, IDBITS))) {
qemu_log_mask(LOG_GUEST_ERROR,
"ITS MAPD: invalid device table attributes "
"devid %d or size %d\n", devid, size);
/*
* in this implementation, in case of error
* we ignore this command and move onto the next
* command in the queue
*/
} else {
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
result = update_dte(s, devid, valid, size, itt_addr) ? CMD_CONTINUE : CMD_STALL;
}
return result;
}
/*
* Current implementation blocks until all
* commands are processed
*/
static void process_cmdq(GICv3ITSState *s)
{
uint32_t wr_offset = 0;
uint32_t rd_offset = 0;
uint32_t cq_offset = 0;
uint64_t data;
AddressSpace *as = &s->gicv3->dma_as;
MemTxResult res = MEMTX_OK;
uint8_t cmd;
int i;
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
return;
}
wr_offset = FIELD_EX64(s->cwriter, GITS_CWRITER, OFFSET);
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if (wr_offset >= s->cq.num_entries) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid write offset "
"%d\n", __func__, wr_offset);
return;
}
rd_offset = FIELD_EX64(s->creadr, GITS_CREADR, OFFSET);
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
if (rd_offset >= s->cq.num_entries) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid read offset "
"%d\n", __func__, rd_offset);
return;
}
while (wr_offset != rd_offset) {
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
ItsCmdResult result = CMD_CONTINUE;
cq_offset = (rd_offset * GITS_CMDQ_ENTRY_SIZE);
data = address_space_ldq_le(as, s->cq.base_addr + cq_offset,
MEMTXATTRS_UNSPECIFIED, &res);
if (res != MEMTX_OK) {
s->creadr = FIELD_DP64(s->creadr, GITS_CREADR, STALLED, 1);
qemu_log_mask(LOG_GUEST_ERROR,
"%s: could not read command at 0x%" PRIx64 "\n",
__func__, s->cq.base_addr + cq_offset);
break;
}
cmd = (data & CMD_MASK);
switch (cmd) {
case GITS_CMD_INT:
result = process_its_cmd(s, data, cq_offset, INTERRUPT);
break;
case GITS_CMD_CLEAR:
result = process_its_cmd(s, data, cq_offset, CLEAR);
break;
case GITS_CMD_SYNC:
/*
* Current implementation makes a blocking synchronous call
* for every command issued earlier, hence the internal state
* is already consistent by the time SYNC command is executed.
* Hence no further processing is required for SYNC command.
*/
break;
case GITS_CMD_MAPD:
result = process_mapd(s, data, cq_offset);
break;
case GITS_CMD_MAPC:
result = process_mapc(s, cq_offset);
break;
case GITS_CMD_MAPTI:
result = process_mapti(s, data, cq_offset, false);
break;
case GITS_CMD_MAPI:
result = process_mapti(s, data, cq_offset, true);
break;
case GITS_CMD_DISCARD:
result = process_its_cmd(s, data, cq_offset, DISCARD);
break;
case GITS_CMD_INV:
case GITS_CMD_INVALL:
/*
* Current implementation doesn't cache any ITS tables,
* but the calculated lpi priority information. We only
* need to trigger lpi priority re-calculation to be in
* sync with LPI config table or pending table changes.
*/
for (i = 0; i < s->gicv3->num_cpu; i++) {
gicv3_redist_update_lpi(&s->gicv3->cpu[i]);
}
break;
default:
break;
}
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
if (result == CMD_CONTINUE) {
rd_offset++;
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
rd_offset %= s->cq.num_entries;
s->creadr = FIELD_DP64(s->creadr, GITS_CREADR, OFFSET, rd_offset);
} else {
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
/* CMD_STALL */
s->creadr = FIELD_DP64(s->creadr, GITS_CREADR, STALLED, 1);
qemu_log_mask(LOG_GUEST_ERROR,
hw/intc/arm_gicv3_its: Use enum for return value of process_* functions When an ITS detects an error in a command, it has an implementation-defined (CONSTRAINED UNPREDICTABLE) choice of whether to ignore the command, proceeding to the next one in the queue, or to stall the ITS command queue, processing nothing further. The behaviour required when the read of the command packet from memory fails is less clearly documented, but the same set of choices as for command errors seem reasonable. The intention of the QEMU implementation, as documented in the comments, is that if we encounter a memory error reading the command packet or one of the various data tables then we should stall, but for command parameter errors we should ignore the queue and continue. However, we don't actually do this. To get the desired behaviour, the various process_* functions need to return true to cause process_cmdq() to advance to the next command and keep processing, and false to stall command processing. What they mostly do is return false for any kind of error. To make the code clearer, replace the 'bool' return from the process_ functions with an enum which may be either CMD_STALL or CMD_CONTINUE. In this commit no behaviour changes; in subsequent commits we will adjust the error-return paths for the process_ functions one by one. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20220111171048.3545974-6-peter.maydell@linaro.org
2022-01-11 18:10:40 +01:00
"%s: 0x%x cmd processing failed, stalling\n",
__func__, cmd);
break;
}
}
}
/*
* This function extracts the ITS Device and Collection table specific
* parameters (like base_addr, size etc) from GITS_BASER register.
* It is called during ITS enable and also during post_load migration
*/
static void extract_table_params(GICv3ITSState *s)
{
uint16_t num_pages = 0;
uint8_t page_sz_type;
uint8_t type;
uint32_t page_sz = 0;
uint64_t value;
for (int i = 0; i < 8; i++) {
TableDesc *td;
int idbits;
value = s->baser[i];
if (!value) {
continue;
}
page_sz_type = FIELD_EX64(value, GITS_BASER, PAGESIZE);
switch (page_sz_type) {
case 0:
page_sz = GITS_PAGE_SIZE_4K;
break;
case 1:
page_sz = GITS_PAGE_SIZE_16K;
break;
case 2:
case 3:
page_sz = GITS_PAGE_SIZE_64K;
break;
default:
g_assert_not_reached();
}
num_pages = FIELD_EX64(value, GITS_BASER, SIZE) + 1;
type = FIELD_EX64(value, GITS_BASER, TYPE);
switch (type) {
case GITS_BASER_TYPE_DEVICE:
td = &s->dt;
idbits = FIELD_EX64(s->typer, GITS_TYPER, DEVBITS) + 1;
break;
case GITS_BASER_TYPE_COLLECTION:
td = &s->ct;
if (FIELD_EX64(s->typer, GITS_TYPER, CIL)) {
idbits = FIELD_EX64(s->typer, GITS_TYPER, CIDBITS) + 1;
} else {
/* 16-bit CollectionId supported when CIL == 0 */
idbits = 16;
}
break;
default:
/*
* GITS_BASER<n>.TYPE is read-only, so GITS_BASER_RO_MASK
* ensures we will only see type values corresponding to
* the values set up in gicv3_its_reset().
*/
g_assert_not_reached();
}
memset(td, 0, sizeof(*td));
td->valid = FIELD_EX64(value, GITS_BASER, VALID);
/*
* If GITS_BASER<n>.Valid is 0 for any <n> then we will not process
* interrupts. (GITS_TYPER.HCC is 0 for this implementation, so we
* do not have a special case where the GITS_BASER<n>.Valid bit is 0
* for the register corresponding to the Collection table but we
* still have to process interrupts using non-memory-backed
* Collection table entries.)
*/
if (!td->valid) {
continue;
}
td->page_sz = page_sz;
td->indirect = FIELD_EX64(value, GITS_BASER, INDIRECT);
td->entry_sz = FIELD_EX64(value, GITS_BASER, ENTRYSIZE) + 1;
td->base_addr = baser_base_addr(value, page_sz);
if (!td->indirect) {
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
td->num_entries = (num_pages * page_sz) / td->entry_sz;
} else {
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
td->num_entries = (((num_pages * page_sz) /
L1TABLE_ENTRY_SIZE) *
(page_sz / td->entry_sz));
}
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
td->num_ids = 1ULL << idbits;
}
}
static void extract_cmdq_params(GICv3ITSState *s)
{
uint16_t num_pages = 0;
uint64_t value = s->cbaser;
num_pages = FIELD_EX64(value, GITS_CBASER, SIZE) + 1;
memset(&s->cq, 0 , sizeof(s->cq));
s->cq.valid = FIELD_EX64(value, GITS_CBASER, VALID);
if (s->cq.valid) {
hw/intc/arm_gicv3_its: Fix various off-by-one errors The ITS code has to check whether various parameters passed in commands are in-bounds, where the limit is defined in terms of the number of bits that are available for the parameter. (For example, the GITS_TYPER.Devbits ID register field specifies the number of DeviceID bits minus 1, and device IDs passed in the MAPTI and MAPD command packets must fit in that many bits.) Currently we have off-by-one bugs in many of these bounds checks. The typical problem is that we define a max_foo as 1 << n. In the Devbits example, we set s->dt.max_ids = 1UL << (GITS_TYPER.Devbits + 1). However later when we do the bounds check we write if (devid > s->dt.max_ids) { /* command error */ } which incorrectly permits a devid of 1 << n. These bugs will not cause QEMU crashes because the ID values being checked are only used for accesses into tables held in guest memory which we access with address_space_*() functions, but they are incorrect behaviour of our emulation. Fix them by standardizing on this pattern: * bounds limits are named num_foos and are the 2^n value (equal to the number of valid foo values) * bounds checks are either if (fooid < num_foos) { good } or if (fooid >= num_foos) { bad } In this commit we fix the handling of the number of IDs in the device table and the collection table, and the number of commands that will fit in the command queue. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
2022-01-07 18:08:00 +01:00
s->cq.num_entries = (num_pages * GITS_PAGE_SIZE_4K) /
GITS_CMDQ_ENTRY_SIZE;
s->cq.base_addr = FIELD_EX64(value, GITS_CBASER, PHYADDR);
s->cq.base_addr <<= R_GITS_CBASER_PHYADDR_SHIFT;
}
}
static MemTxResult gicv3_its_translation_write(void *opaque, hwaddr offset,
uint64_t data, unsigned size,
MemTxAttrs attrs)
{
GICv3ITSState *s = (GICv3ITSState *)opaque;
bool result = true;
uint32_t devid = 0;
switch (offset) {
case GITS_TRANSLATER:
if (s->ctlr & R_GITS_CTLR_ENABLED_MASK) {
devid = attrs.requester_id;
result = process_its_cmd(s, data, devid, NONE);
}
break;
default:
break;
}
if (result) {
return MEMTX_OK;
} else {
return MEMTX_ERROR;
}
}
static bool its_writel(GICv3ITSState *s, hwaddr offset,
uint64_t value, MemTxAttrs attrs)
{
bool result = true;
int index;
switch (offset) {
case GITS_CTLR:
if (value & R_GITS_CTLR_ENABLED_MASK) {
s->ctlr |= R_GITS_CTLR_ENABLED_MASK;
extract_table_params(s);
extract_cmdq_params(s);
s->creadr = 0;
process_cmdq(s);
} else {
s->ctlr &= ~R_GITS_CTLR_ENABLED_MASK;
}
break;
case GITS_CBASER:
/*
* IMPDEF choice:- GITS_CBASER register becomes RO if ITS is
* already enabled
*/
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
s->cbaser = deposit64(s->cbaser, 0, 32, value);
s->creadr = 0;
s->cwriter = s->creadr;
}
break;
case GITS_CBASER + 4:
/*
* IMPDEF choice:- GITS_CBASER register becomes RO if ITS is
* already enabled
*/
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
s->cbaser = deposit64(s->cbaser, 32, 32, value);
s->creadr = 0;
s->cwriter = s->creadr;
}
break;
case GITS_CWRITER:
s->cwriter = deposit64(s->cwriter, 0, 32,
(value & ~R_GITS_CWRITER_RETRY_MASK));
if (s->cwriter != s->creadr) {
process_cmdq(s);
}
break;
case GITS_CWRITER + 4:
s->cwriter = deposit64(s->cwriter, 32, 32, value);
break;
case GITS_CREADR:
if (s->gicv3->gicd_ctlr & GICD_CTLR_DS) {
s->creadr = deposit64(s->creadr, 0, 32,
(value & ~R_GITS_CREADR_STALLED_MASK));
} else {
/* RO register, ignore the write */
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write to RO register at offset "
TARGET_FMT_plx "\n", __func__, offset);
}
break;
case GITS_CREADR + 4:
if (s->gicv3->gicd_ctlr & GICD_CTLR_DS) {
s->creadr = deposit64(s->creadr, 32, 32, value);
} else {
/* RO register, ignore the write */
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write to RO register at offset "
TARGET_FMT_plx "\n", __func__, offset);
}
break;
case GITS_BASER ... GITS_BASER + 0x3f:
/*
* IMPDEF choice:- GITS_BASERn register becomes RO if ITS is
* already enabled
*/
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
index = (offset - GITS_BASER) / 8;
if (offset & 7) {
value <<= 32;
value &= ~GITS_BASER_RO_MASK;
s->baser[index] &= GITS_BASER_RO_MASK | MAKE_64BIT_MASK(0, 32);
s->baser[index] |= value;
} else {
value &= ~GITS_BASER_RO_MASK;
s->baser[index] &= GITS_BASER_RO_MASK | MAKE_64BIT_MASK(32, 32);
s->baser[index] |= value;
}
}
break;
case GITS_IIDR:
case GITS_IDREGS ... GITS_IDREGS + 0x2f:
/* RO registers, ignore the write */
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write to RO register at offset "
TARGET_FMT_plx "\n", __func__, offset);
break;
default:
result = false;
break;
}
return result;
}
static bool its_readl(GICv3ITSState *s, hwaddr offset,
uint64_t *data, MemTxAttrs attrs)
{
bool result = true;
int index;
switch (offset) {
case GITS_CTLR:
*data = s->ctlr;
break;
case GITS_IIDR:
*data = gicv3_iidr();
break;
case GITS_IDREGS ... GITS_IDREGS + 0x2f:
/* ID registers */
*data = gicv3_idreg(offset - GITS_IDREGS);
break;
case GITS_TYPER:
*data = extract64(s->typer, 0, 32);
break;
case GITS_TYPER + 4:
*data = extract64(s->typer, 32, 32);
break;
case GITS_CBASER:
*data = extract64(s->cbaser, 0, 32);
break;
case GITS_CBASER + 4:
*data = extract64(s->cbaser, 32, 32);
break;
case GITS_CREADR:
*data = extract64(s->creadr, 0, 32);
break;
case GITS_CREADR + 4:
*data = extract64(s->creadr, 32, 32);
break;
case GITS_CWRITER:
*data = extract64(s->cwriter, 0, 32);
break;
case GITS_CWRITER + 4:
*data = extract64(s->cwriter, 32, 32);
break;
case GITS_BASER ... GITS_BASER + 0x3f:
index = (offset - GITS_BASER) / 8;
if (offset & 7) {
*data = extract64(s->baser[index], 32, 32);
} else {
*data = extract64(s->baser[index], 0, 32);
}
break;
default:
result = false;
break;
}
return result;
}
static bool its_writell(GICv3ITSState *s, hwaddr offset,
uint64_t value, MemTxAttrs attrs)
{
bool result = true;
int index;
switch (offset) {
case GITS_BASER ... GITS_BASER + 0x3f:
/*
* IMPDEF choice:- GITS_BASERn register becomes RO if ITS is
* already enabled
*/
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
index = (offset - GITS_BASER) / 8;
s->baser[index] &= GITS_BASER_RO_MASK;
s->baser[index] |= (value & ~GITS_BASER_RO_MASK);
}
break;
case GITS_CBASER:
/*
* IMPDEF choice:- GITS_CBASER register becomes RO if ITS is
* already enabled
*/
if (!(s->ctlr & R_GITS_CTLR_ENABLED_MASK)) {
s->cbaser = value;
s->creadr = 0;
s->cwriter = s->creadr;
}
break;
case GITS_CWRITER:
s->cwriter = value & ~R_GITS_CWRITER_RETRY_MASK;
if (s->cwriter != s->creadr) {
process_cmdq(s);
}
break;
case GITS_CREADR:
if (s->gicv3->gicd_ctlr & GICD_CTLR_DS) {
s->creadr = value & ~R_GITS_CREADR_STALLED_MASK;
} else {
/* RO register, ignore the write */
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write to RO register at offset "
TARGET_FMT_plx "\n", __func__, offset);
}
break;
case GITS_TYPER:
/* RO registers, ignore the write */
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write to RO register at offset "
TARGET_FMT_plx "\n", __func__, offset);
break;
default:
result = false;
break;
}
return result;
}
static bool its_readll(GICv3ITSState *s, hwaddr offset,
uint64_t *data, MemTxAttrs attrs)
{
bool result = true;
int index;
switch (offset) {
case GITS_TYPER:
*data = s->typer;
break;
case GITS_BASER ... GITS_BASER + 0x3f:
index = (offset - GITS_BASER) / 8;
*data = s->baser[index];
break;
case GITS_CBASER:
*data = s->cbaser;
break;
case GITS_CREADR:
*data = s->creadr;
break;
case GITS_CWRITER:
*data = s->cwriter;
break;
default:
result = false;
break;
}
return result;
}
static MemTxResult gicv3_its_read(void *opaque, hwaddr offset, uint64_t *data,
unsigned size, MemTxAttrs attrs)
{
GICv3ITSState *s = (GICv3ITSState *)opaque;
bool result;
switch (size) {
case 4:
result = its_readl(s, offset, data, attrs);
break;
case 8:
result = its_readll(s, offset, data, attrs);
break;
default:
result = false;
break;
}
if (!result) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest read at offset " TARGET_FMT_plx
"size %u\n", __func__, offset, size);
/*
* The spec requires that reserved registers are RAZ/WI;
* so use false returns from leaf functions as a way to
* trigger the guest-error logging but don't return it to
* the caller, or we'll cause a spurious guest data abort.
*/
*data = 0;
}
return MEMTX_OK;
}
static MemTxResult gicv3_its_write(void *opaque, hwaddr offset, uint64_t data,
unsigned size, MemTxAttrs attrs)
{
GICv3ITSState *s = (GICv3ITSState *)opaque;
bool result;
switch (size) {
case 4:
result = its_writel(s, offset, data, attrs);
break;
case 8:
result = its_writell(s, offset, data, attrs);
break;
default:
result = false;
break;
}
if (!result) {
qemu_log_mask(LOG_GUEST_ERROR,
"%s: invalid guest write at offset " TARGET_FMT_plx
"size %u\n", __func__, offset, size);
/*
* The spec requires that reserved registers are RAZ/WI;
* so use false returns from leaf functions as a way to
* trigger the guest-error logging but don't return it to
* the caller, or we'll cause a spurious guest data abort.
*/
}
return MEMTX_OK;
}
static const MemoryRegionOps gicv3_its_control_ops = {
.read_with_attrs = gicv3_its_read,
.write_with_attrs = gicv3_its_write,
.valid.min_access_size = 4,
.valid.max_access_size = 8,
.impl.min_access_size = 4,
.impl.max_access_size = 8,
.endianness = DEVICE_NATIVE_ENDIAN,
};
static const MemoryRegionOps gicv3_its_translation_ops = {
.write_with_attrs = gicv3_its_translation_write,
.valid.min_access_size = 2,
.valid.max_access_size = 4,
.impl.min_access_size = 2,
.impl.max_access_size = 4,
.endianness = DEVICE_NATIVE_ENDIAN,
};
static void gicv3_arm_its_realize(DeviceState *dev, Error **errp)
{
GICv3ITSState *s = ARM_GICV3_ITS_COMMON(dev);
int i;
for (i = 0; i < s->gicv3->num_cpu; i++) {
if (!(s->gicv3->cpu[i].gicr_typer & GICR_TYPER_PLPIS)) {
error_setg(errp, "Physical LPI not supported by CPU %d", i);
return;
}
}
gicv3_its_init_mmio(s, &gicv3_its_control_ops, &gicv3_its_translation_ops);
address_space_init(&s->gicv3->dma_as, s->gicv3->dma,
"gicv3-its-sysmem");
/* set the ITS default features supported */
hw/intc/arm_gicv3_its: Don't misuse GITS_TYPE_PHYSICAL define The GITS_TYPE_PHYSICAL define is the value we set the GITS_TYPER.Physical field to -- this is 1 to indicate that we support physical LPIs. (Support for virtual LPIs is the GITS_TYPER.Virtual field.) We also use this define as the *value* that we write into an interrupt translation table entry's INTTYPE field, which should be 1 for a physical interrupt and 0 for a virtual interrupt. Finally, we use it as a *mask* when we read the interrupt translation table entry INTTYPE field. Untangle this confusion: define an ITE_INTTYPE_VIRTUAL and ITE_INTTYPE_PHYSICAL to be the valid values of the ITE INTTYPE field, and replace the ad-hoc collection of ITE_ENTRY_* defines with use of the FIELD() macro to define the fields of an ITE and the FIELD_EX64() and FIELD_DP64() macros to read and write them. We use ITE in the new setup, rather than ITE_ENTRY, because ITE stands for "Interrupt translation entry" and so the extra "entry" would be redundant. We take the opportunity to correct the name of the field that holds the GICv4 'doorbell' interrupt ID (this is always the value 1023 in a GICv3, which is why we were calling it the 'spurious' field). The GITS_TYPE_PHYSICAL define is then used in only one place, where we set the initial GITS_TYPER value. Since GITS_TYPER.Physical is essentially a boolean, hiding the '1' value behind a macro is more confusing than helpful, so expand out the macro there and remove the define entirely. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
2022-01-07 18:07:59 +01:00
s->typer = FIELD_DP64(s->typer, GITS_TYPER, PHYSICAL, 1);
s->typer = FIELD_DP64(s->typer, GITS_TYPER, ITT_ENTRY_SIZE,
ITS_ITT_ENTRY_SIZE - 1);
s->typer = FIELD_DP64(s->typer, GITS_TYPER, IDBITS, ITS_IDBITS);
s->typer = FIELD_DP64(s->typer, GITS_TYPER, DEVBITS, ITS_DEVBITS);
s->typer = FIELD_DP64(s->typer, GITS_TYPER, CIL, 1);
s->typer = FIELD_DP64(s->typer, GITS_TYPER, CIDBITS, ITS_CIDBITS);
}
static void gicv3_its_reset(DeviceState *dev)
{
GICv3ITSState *s = ARM_GICV3_ITS_COMMON(dev);
GICv3ITSClass *c = ARM_GICV3_ITS_GET_CLASS(s);
c->parent_reset(dev);
/* Quiescent bit reset to 1 */
s->ctlr = FIELD_DP32(s->ctlr, GITS_CTLR, QUIESCENT, 1);
/*
* setting GITS_BASER0.Type = 0b001 (Device)
* GITS_BASER1.Type = 0b100 (Collection Table)
* GITS_BASER<n>.Type,where n = 3 to 7 are 0b00 (Unimplemented)
* GITS_BASER<0,1>.Page_Size = 64KB
* and default translation table entry size to 16 bytes
*/
s->baser[0] = FIELD_DP64(s->baser[0], GITS_BASER, TYPE,
GITS_BASER_TYPE_DEVICE);
s->baser[0] = FIELD_DP64(s->baser[0], GITS_BASER, PAGESIZE,
GITS_BASER_PAGESIZE_64K);
s->baser[0] = FIELD_DP64(s->baser[0], GITS_BASER, ENTRYSIZE,
GITS_DTE_SIZE - 1);
s->baser[1] = FIELD_DP64(s->baser[1], GITS_BASER, TYPE,
GITS_BASER_TYPE_COLLECTION);
s->baser[1] = FIELD_DP64(s->baser[1], GITS_BASER, PAGESIZE,
GITS_BASER_PAGESIZE_64K);
s->baser[1] = FIELD_DP64(s->baser[1], GITS_BASER, ENTRYSIZE,
GITS_CTE_SIZE - 1);
}
static void gicv3_its_post_load(GICv3ITSState *s)
{
if (s->ctlr & R_GITS_CTLR_ENABLED_MASK) {
extract_table_params(s);
extract_cmdq_params(s);
}
}
static Property gicv3_its_props[] = {
DEFINE_PROP_LINK("parent-gicv3", GICv3ITSState, gicv3, "arm-gicv3",
GICv3State *),
DEFINE_PROP_END_OF_LIST(),
};
static void gicv3_its_class_init(ObjectClass *klass, void *data)
{
DeviceClass *dc = DEVICE_CLASS(klass);
GICv3ITSClass *ic = ARM_GICV3_ITS_CLASS(klass);
GICv3ITSCommonClass *icc = ARM_GICV3_ITS_COMMON_CLASS(klass);
dc->realize = gicv3_arm_its_realize;
device_class_set_props(dc, gicv3_its_props);
device_class_set_parent_reset(dc, gicv3_its_reset, &ic->parent_reset);
icc->post_load = gicv3_its_post_load;
}
static const TypeInfo gicv3_its_info = {
.name = TYPE_ARM_GICV3_ITS,
.parent = TYPE_ARM_GICV3_ITS_COMMON,
.instance_size = sizeof(GICv3ITSState),
.class_init = gicv3_its_class_init,
.class_size = sizeof(GICv3ITSClass),
};
static void gicv3_its_register_types(void)
{
type_register_static(&gicv3_its_info);
}
type_init(gicv3_its_register_types)