sceNetEpollCreate fails with EINVAL for a null name or non-zero flags,
sceNetResolverCreate and sceNetGetMacAddress for non-zero flags, and
sceNetResolverStartNtoa, before any other check, for a negative timeout
or retry count or a flag bit outside 0x10001, as prosper reads the
libSceNet module. The two accepted resolver flag bits are still ignored,
which the debt file now records.
An image_store with DIM 2D_ARRAY through a 1D T# had no runtime image
mode, so the dispatch stopped with "image descriptor is incompatible
with the static runtime image interface". #2348 added a 1D mode for
DIM 2D stores only.
Write-only 2D-array storage images now get the same 1D mode as the 2D
case. On RDNA2 (gfx1036, hardware oracle) such a store writes texel x
and ignores y when the slice is 0, and writes nothing for any other
slice, with DEPTH 0 or 3, so the store is skipped unless the slice is 0.
New execution test agc_driver_image_store_line_through_plane_array.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The RDNA program decode sweeps linearly up to the last s_endpgm past the
furthest direct branch target, so it also decoded words that no path
reaches and threw on any it could not decode. A word directly after
s_branch, s_endpgm or s_code_end that fails to decode is now skipped;
the decode still throws that word's error when a direct branch or
s_call_b64 targets it, and a failure on a word reachable by fall-through
is unchanged. The control-flow graph starts the block after s_branch and
s_endpgm at the next decoded instruction instead of the next word.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
On PS5 firmware 13.60 sceRtcSetTime_t(NULL, -1) returns
SCE_RTC_ERROR_INVALID_VALUE and sceRtcSetTime_t(NULL, 0)
SCE_RTC_ERROR_INVALID_POINTER: negative seconds are rejected before a
null destination under the later SDK rules. The other measured
GetDayOfWeek, SetTime_t and FormatRFC3339 results already matched.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Apply the firmware 13.42 module-probe report in PR #9506: all four DNS calls check initialization first; getters return EINVAL for null outputs or nonzero flags; setters reject flags before reading guest memory. Keep the checked input copy and document EFAULT as host safety, not firmware behavior. Use the current public Net.hpp declarations in the regression.
Rebased onto upstream e7e3e9c198. Ubuntu 24.04, GCC 13.3, Release: actual libc/libkernel/libSceNet libraries and both focused test executables built. guest_sce_net passes. A separate harness linked against those unmodified libraries passes initialization, errno, invalid flags, unreadable-input flag precedence, null clearing and unchanged-output checks. The full guest_net_dns_info test is blocked here: process_vm_readv is denied with EPERM by the execution environment. No full DNS regression pass is claimed. Public-header audit, conventions, test registration and whitespace checks pass.
The resolver still uses host DNS; IPv6 host-server discovery and firmware server-count behavior remain debt. Firmware evidence: https://github.com/boykopovar/AnyPS5/pull/9506#issuecomment-6390574180
AI-assisted: yes
On PS5 firmware 13.60 setpriority(PRIO_PROCESS, 0, v) returns 0 for -25,
-5, 5 and 25, and getpriority keeps returning 0. setpriority keeps its
target checks and returns 0 without storing the value.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Walk each macOS function import before the guest entry and pass it through TraceImport_nid_no_patch, as Linux and Windows already do. An x86-64 libc build records the call and prints it on SIGABRT.
Both exports threw and were declared without parameters. They follow
LibreSSL 2.5.4, unchanged in 2.7: X509_LOOKUP_file returns the static
80-byte file method, and X509_STORE_add_lookup returns the store's
lookup for a method, or mallocs a 32-byte lookup owned by the store and
pushes it on the store's methods, growing the stack the way sk_insert
does. X509_STORE_free now frees the store's lookups.
The file method's ctrl, which loads PEM files, throws: there is no
certificate model yet. Callbacks of guest-supplied methods throw
instead of being called.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
Both exports threw and were declared without parameters. They follow
LibreSSL 2.5.4, unchanged in 2.7: set_hostflags stores the host
matching flags in the verify parameters' identity, and set1_ip_asc
parses an IPv4 or IPv6 address the way a2i_ipadd does, copies the 4 or
16 bytes into a guest malloc block that replaces the previous address,
and returns 0 without touching it for text the parser rejects. IPv4
parts are read with the guest sscanf, as the module would.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
The three exports threw and were declared without parameters. They now
follow LibreSSL 2.5.4, the version the firmware 13.40 measurements
identify: X509_STORE_new allocates the 136-byte store with an empty
object cache, an empty lookup method stack and 64-byte verify
parameters at depth -1, X509_STORE_free drops a reference and frees the
store with its stacks, verify parameters and their host names, peer
name, email and IP, and X509_STORE_set_flags ORs the flags into the
verify parameters and adds the policy check for any policy flag.
No export adds certificates, CRLs or lookup methods yet, so freeing a
store that holds them, or verify parameters holding policy objects,
throws instead of freeing objects with no model.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A null batch address is refused whatever the size, so the accepted batch
of zero bytes is one at a real address, which is what TestConsoleEmptyBatch
starts.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
A probe on a PS4 decoded exactly one frame and reported 0 for an ATRAC9
run job without RUN_MULTIPLE_FRAMES. Jobs read from a PS4 batch buffer
now do the same. Jobs from the PS5 builders keep decoding their whole
input, which FMOD in PPSA29141 relies on.
Co-Authored-By: Claude Opus 5.5 <[email protected]>
sceAjmBatchStartBuffer let a null batch of zero bytes through its
parameter check and formed the end pointer from it. A null batch now
returns INVALID_PARAMETER whatever the size, as the PS5 builders'
sceAjmBatchInitialize does; a batch of zero bytes still starts, runs
nothing and completes at once. A test covers both; neither was measured
on a console, which the debt entry says.
Co-Authored-By: Claude Fable 5.1 <[email protected]>
AZLisme