nikola/llama_cpp
1
0
Fork 0
Code Issues Pull Requests Actions 4 Packages Projects Releases Wiki Activity

app : enable self-update only when built with llama-install.sh (#24754)

Browse Source 
Signed-off-by: Adrien Gallouët <angt@huggingface.co>
This commit is contained in:
Adrien Gallouët
2026-06-18 09:57:59 +02:00
committed by GitHub
parent 32e806b9c1
commit 6ec59ddaea
1 changed files with 26 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

39
app/llama.cpp
View File

@@ -20,16 +20,21 @@ int llama_fit_params(int argc, char ** argv);
int llama_quantize(int argc, char ** argv);
int llama_perplexity(int argc, char ** argv);
// hands the update over to the install script, which downloads and swaps the binary
// Self-update is only supported for binaries built with llama-install.sh
static int llama_update(int argc, char ** argv) {
(void) argc;
(void) argv;
#ifdef LLAMA_INSTALL_BUILD
#if defined(_WIN32)
return system("powershell -NoProfile -ExecutionPolicy Bypass -Command \"irm https://llama.app/install.ps1 | iex\"");
#else
return system("curl -fsSL https://llama.app/install.sh | sh");
#endif
#else
printf("Updates are available only when installed from https://llama.app\n");
return 1;
#endif
}
static const char * progname;
@@ -46,21 +51,29 @@ struct command {
int (*func)(int, char **);
};
#ifdef LLAMA_INSTALL_BUILD
#define UPDATE_HIDDEN false
#else
#define UPDATE_HIDDEN true
#endif
static const command cmds[] = {
{"serve", "HTTP API server", {"server"}, false, llama_server },
{"cli", "Command-line interactive interface", {"client"}, false, llama_cli },
{"update", "Update llama to the latest release", {}, false, llama_update },
{"completion", "Text completion", {"complete"}, true, llama_completion },
{"bench", "Benchmark prompt processing and text generation", {}, true, llama_bench },
{"batched-bench", "Benchmark batched decoding performance", {}, true, llama_batched_bench},
{"fit-params", "Compute parameters to fit a model in device memory", {}, true, llama_fit_params },
{"quantize", "Quantize a model", {}, true, llama_quantize },
{"perplexity", "Compute model perplexity and KL divergence", {}, true, llama_perplexity },
{"version", "Show version", {}, false, version },
{"licenses", "Show third-party licenses", {"credits"}, false, licenses },
{"help", "Show available commands", {}, false, help },
{"serve", "HTTP API server", {"server"}, false, llama_server },
{"cli", "Command-line interactive interface", {"client"}, false, llama_cli },
{"update", "Update llama to the latest release", {}, UPDATE_HIDDEN, llama_update },
{"completion", "Text completion", {"complete"}, true, llama_completion },
{"bench", "Benchmark prompt processing and text generation", {}, true, llama_bench },
{"batched-bench", "Benchmark batched decoding performance", {}, true, llama_batched_bench},
{"fit-params", "Compute parameters to fit a model in device memory", {}, true, llama_fit_params },
{"quantize", "Quantize a model", {}, true, llama_quantize },
{"perplexity", "Compute model perplexity and KL divergence", {}, true, llama_perplexity },
{"version", "Show version", {}, false, version },
{"licenses", "Show third-party licenses", {"credits"}, false, licenses },
{"help", "Show available commands", {}, false, help },
};
#undef UPDATE_HIDDEN
static int version(int argc, char ** argv) {
printf("%s\n", llama_build_info());
return 0;